Does a High Minutes Average Cause Shooting Fatigue?

   Welcome back! In today's article, I wanted to look at the effects of playing nearly every minute of each game in a college season. Fortunately, there were several players who fit this criteria from last season, but we'll first narrow the focus to just the top five players with the highest minutes per game averages. All of these players played more than 38 minutes each night. With less than two minutes of rest each game, I was wondering if these players' fatigue would show itself in their shooting percentages. I thought that after around 30 minutes of action, exhaustion would set in and we'd see their shooting percentages dip a bit as a result. I was specifically thinking the 30 minute mark because the halftime rest is wearing off, and yet there's still 10 more minutes left to play (assuming no overtime). I also figured that said exhaustion would become less relevant in the final five minutes and adrenaline would take over as the players tried to mount comebacks or maintain a lead for the win. Thus, I was anticipating a noticeable decline in shooting percentage somewhere around minutes 30-35 and then an increase in the final 5 minutes.

How?

    As mentioned above, I decided to test the idea on the top five players with the highest minutes per game averages last season. These ended up being George Tinsley (Binghamton), Kyle Lofton (St. Bonaventure), Nate Darling (Delaware - now in the NBA Draft), Malik Johnson (Canisius), and Jason Preston (Ohio). For each player, I created a graph of their cumulative shooting percentage for the previous five minutes. Basically, this means that you can look at say minute 25 and see the player's shooting percentage across the previous five minutes (minutes 20-25) for all games combined. It lets us pick any given minute and see how a player shot for the previous five minute interval across the entire season. Here's what this looks like for Binghamton's George Tinsley:

    The first thing to emphasize is that the graph looks weird up until the 5 minute mark. This is because we're supposed to be looking at shooting percentage for the previous five minutes, but there isn't a full five minute interval to look back on yet. For instance, if a player went 0/1 in the first minute across all games, the cumulative shooting percentage would be 0% at the 1 minute mark on the graph. Similarly, if he went 1/1, the average would be 100%. The sample size of shots isn't maximized until the 5 minute mark and the graph yields relatively extreme results until then. Hence, we have to wait until this point before we get a complete sample of shots to begin analysis.

    

    Another important aspect to point out is the red dotted line that I've included which signifies a player's average shooting percentage for the season. For the example above with George Tinsley, it's roughly 40.7%. By including the average, we can see streakiness in a shooter. In Tinsley's case, we see that his peak cumulative shooting percentage occurred around minute 32. This means that in the five minute interval between minute 27 and minute 32, Tinsley shot 60% - nearly 20% above his season-long average shooting percentage. Conversely, his low was at minute 39 and so he shot below 25% between minutes 34 and 39 which was 15% below his average shooting percentage. 

Graph Results

    After going through the intricacies of what Tinsley's graph is really showing, I'll just dump the rest of the players' graphs below without any more comments. Then we can dive into whether these graphs support my initial idea.

Is the Initial Idea Proven?

    My original theory was that these high minutes players would start feeling the fatigue setting in around 30-35 minutes into the game. This would cause shooting percentage to drop which then would be followed by a recovery in the final 5 minutes. If this were the case, we'd expect to see a low on the above graphs at 35 minutes (since this would represent the cumulative shooting percentage for minutes 30-35). We then would expect the cumulative shooting percentage to gradually increase through minute 40. Unfortunately, this trend didn't really appear in the above five graphs. Kyle Lofton and Jason Preston were the only players who had a cumulative shooting percentage lower than their overall average shooting percentage at minute 35. If we look for the impact of fatigue slightly earlier in the game, there does appear to be some evidence that fatigue sets in between 27.5-32.5 minutes as seen on the cumulative average graphs as a low at minute 32.5. Darling and Preston both have pretty clear lows around this minute mark. Having only 2/5 players support this trend isn't strong evidence at all, but just so that we don't miss out on a key trend due to our small sample, we can increase it to ten players. To achieve this, we'll look at another five players who had the next highest minutes per game averages behind our original five. These players include Jordan Ford (St. Mary's), Michael Ertel (Louisiana-Monroe), Jermaine Marrow (Hampton), Mike Smith (Columbia), and Colbey Ross (Pepperdine). Their graphs are below:

Additional Graphs

Impact of a Larger Sample

    After adding these five new graphs to our original sample of five, it doesn't appear that minute 32.5 is a common point of fatigue for these high usage players. Ertel, Marrow, and Ross all have a cumulative shooting percentage lower than their overall averages at minute 32.5, but Ertel and Marrow's cumulative shooting percentages continued to decline to significantly further lows after this point. Also, the other two players (Ford and Smith) had above average cumulative shooting percentages at this point. Ultimately, there just isn't a clear enough trend amongst these players to label a specific time point or even relatively specific range of time where fatigue seemingly causes all of these players' shooting percentages to be lower than normal. 

    Although these ten players had essentially the same minutes distribution since they were only off the court for about two minutes per game, the effects of fatigue on shooting seemingly were not the same. This could come down to a variety of factors including differing levels of player conditioning (one player may be impacted less or at different times by fatigue compared to another player). Another possibility is that the roughly two minutes of rest per game that these players got could have had a greater impact than the negligible one that I was assuming. For instance, I originally believed that the 35 minute mark would be a low point for most of these players. However, if a player tended to get the full amount of his minimal rest right before then (as opposed to the rest being diluted in smaller intervals across the game), it could be enough to make his level of fatigue significantly lower than it should be at this point in the game otherwise. This key timing of the small amount of rest might allow the player to shoot better than he would've without it. This would then go against the general player fatigue trend and show itself in a peak on the graph where a lot of other players had a low. 

What Was the Purpose of All of This?

    You might be thinking that this whole article was quite pointless if we didn't end up pinpointing a common area of low cumulative shooting percentage due to fatigue. In this sense, yes, the article failed. However, it was important in seeing that the impact of fatigue is pretty tricky to pick out, even within players who essentially played at the same times (the entire game). This article doesn't prove that fatigue has no universal impact on players' shooting percentages, it just confirms that such an impact cannot be found with this type of analysis. Another key takeaway is that there were a lot of interesting player-specific results from creating these cumulative shooting percentage graphs (See: Jordan Ford's volatile first half shooting compared to his relative consistency in the second half). I'll dive more into these other takeaways from the graphs hopefully in the near future in another article.

    If you've made it this far, thank you so much! I know that this was a bit of a long and complicated one. If you enjoyed it, be sure to check out some of the other articles that I've written on this blog!

Is an Elite ORB% Sustainable?

   Welcome back! In writing a previous article about the top ORB% leaders in the country for the 2019-2020 season, I found that I was very optimistic that the guys on the list would be able to sustain at least similarly high offensive rebounding rates heading into next season. This was especially true for players averaging 20 mpg or less, as I was excited to see what they could do with additional minutes. However, I began to wonder if such optimism was warranted. With more minutes in the next season, we would expect ORB% to decrease slightly. This is because ORB% measures the percentage of available offensive rebounds that a player grabs while he's on the court. Thus, more minutes played means more offensive rebounds a player needs to get in order to maintain the same rate. I was hoping to determine just how big this ORB% drop off is, and if it's small enough for players to still have an elite ORB%. Also, I wanted to see if the drop off was worth the increase in minutes.

    To find out, I looked at the top 50 players in terms of ORB% from the past 4 seasons on Sports Reference. I removed players whose top 50 ORB% season was his last in college and any players who transferred. I then calculated the minutes per game increase or decrease following the elite ORB% season and how much the player's ORB% changed after the top 50 season. For instance, Michigan State's Nick Ward had a top 50 ORB% season in 2017-2018, so I compared his minutes per game and ORB% to the following 2018-2019 season. Here's the results:

    As you can see, the vast majority of players experienced the expected decline in ORB% as their minutes increased. To answer the first question posed in the introduction, most players saw a decline of about 1-6%. While it's a decent decrease, to be in the top 50 in terms of ORB%, these players had an ORB% of somewhere between 13-18% in their standout season. Even a drop of 6% in the next season would mean at worst an ORB% of 7%. This is hardly elite anymore, but isn't completely awful.

    Somewhat surprisingly, ORB% also decreased in players who experienced significant minute reductions. You'd think that a player's ORB% would go up with fewer minutes because they'd have to be locked in on getting offensive rebounds for less time per game to maintain the same ORB%. Perhaps ORB% mostly declines along with minutes due to underlying poor play. For instance, coaches may have been unimpressed during the offseason/beginning few games and decided to cut these players' minutes. In this scenario, the players may have simply been not as good as they were in the prior season which could yield fewer offensive boards and a lower ORB%. 

    While ORB% seems to decline after an elite season for most players, there are a few who buck this trend and preserve or improve their already high ORB% in the following season. Most of such players experienced minimal changes in the amount of minutes that they played. This might suggest that these players were already operating in a role that maximized their offensive rebounding efficiency. By either pure luck or intentional strategy, their coaches opted to keep these players' minutes the same and benefited from another excellent ORB% season. Unfortunately, there are also plenty of players who undermined this pattern as they played roughly the same minutes the following season and saw their ORB% drop, sometimes significantly. Thus, getting a player to maintain an exceptional ORB% isn't as simple as holding his minutes the same.

    Given that we largely see a decline in ORB% in most players who saw more minutes in the next season, it's important to take a look at what's driving this drop. Is it that players are getting fewer offensive rebounds overall in the following season, or is the minute increase greater than a player's offensive rebounding increase? To achieve this, we can color each point of the above graph according to how many more (or fewer) offensive rebounds per game a player got in the next season. Green indicates a large increase and red indicates a large decrease. Here's what the new, color-coded graph looks like:

    As would be expected, pretty much every player who experienced a minutes reduction also grabbed fewer offensive rebounds in the following season. It certainly makes sense as less time on the court means less opportunities to get rebounds. Also, the few standouts who were able to meet or exceed their previously impressive ORB% were actually able to get more offensive rebounds per game as well. These players were especially useful as they maintained their offensive rebounding efficiency while also increasing their volume of offensive rebounds. Most interestingly, we see from the graph that the majority of players who saw increased minutes also increased their rebounds per game. This means that their decrease in ORB% was caused by their minute increase outpacing their offensive rebounding increase. This is important because it suggests that the decrease in ORB% may not be such a bad thing after all. When these players were given expanded minutes, their offensive rebounding didn't suddenly get much worse, it just didn't improve at quite as rapid of a rate.

    It's also important to consider what else these players are doing in their increased minutes. After all, most coaches won't give a guy 5-10 extra minutes per game expecting him to only focus on increasing his offensive rebounding. An area that I thought would be especially important to look at is scoring. If a player is going to play significantly more per game, they'll likely need to chip in with more points. The following graph is the same as the previous two, but is colored by the points per game increase or decrease in the following season. Like the last graph, green indicates the largest scoring increase and red represents the largest drop.

   Mirroring the trend in the previous graph of players who had a minutes reduction getting fewer offensive rebounds per game, these players also scored less. Even players who kept roughly the same minutes and maintained or improved their ORB% varied in terms of if they were able to increase their scoring or not. While these players kept a high rebounding efficiency and increased their rebounding volume, scoring didn't always follow suit. Meanwhile, most players who experienced an increase in minutes also experienced a scoring boost, even if it was just a slight one. Thus, this benefit to the team serves to somewhat counteract the ORB% decline.

    After diving into these three graphs, it's time to return to the second question posed in the introduction regarding if the minutes increase is worth it. The answer to this is that it really depends. The second and third graphs revealed that a decrease in ORB% isn't inherently a bad thing because players can still increase their offensive rebounding volume at the expense of some efficiency while also improving other skills such as scoring. However, the tricky part is determining whether the incremental offensive rebounding and/or scoring increase is worth sacrificing an excellent ORB%. I think it's safe to say that players such as those represented by the two green points furthest to the right on the above graph are cases where it's worth it. An increase of 5-10 points per game is likely more valuable to their respective teams than about 5% of lost ORB% considering that their rebounds per game also increased. It gets more difficult when comparing say a 1 point per game scoring increase to a 3% drop in ORB%. 

    Ignoring team-specific roster situations, it seems as though in general coaches may as well risk worse offensive rebounding efficiency with increased minutes in the hopes of more overall production from their players. Chasing the same elite ORB% efficiency as the previous season by keeping minutes the same appears to be just as risky. Overall, my optimism in the previous article was a little excessive because we've seen that an elite ORB% is only sustainable in rare instances. However, this sustainability doesn't seem to be desirable to aim for in the first place. 

    Thanks so much for reading if you've made it this far! Be sure to check out some of my other articles if you enjoyed this!

Offensive Rebounding Leaders vs. Defensive Rebounding Leaders

    Welcome back! If you missed my previous article comparing offensive rebounding percentage (the percentage of available offensive rebounds a player got while he was in the game) to offensive ratings (the number of points that a player produces for every 100 possessions) for the top 20 players in terms of ORB%, you can find it here. If you missed the similar article comparing defensive rebounding percentage (the opposite of ORB%) to defensive ratings (the number of points that a player allows for every 100 possessions), you can read that one here. This article combines the data from those two articles, so it might be helpful to skim them at least for reference.

    The goal of this article is to compare ORB% to DRB%. This will help us to figure out which players are specialists on the offensive glass only, defensive glass only, and who are all-around dominant rebounders. We'll look at both the ORB% and DRB% for the top 20 players in terms of ORB% and the top 20 players in terms of DRB%. Essentially, we're combining the two sets of players from the previous two articles. Both the ORB% and DRB% data comes from Fox Sports. Any additional data comes from Fox Sports or Sports Reference.

Here's what the graph looks like comparing ORB% and DRB%:

As always, here's the complete list of players for reference:

Oscar Tshiebwe - West Virginia

Jordan Minor - Merrimack

Ed Croswell - LaSalle (now Providence)

Kevin Marfo - Quinnipiac (now Texas A&M)

Trevion Williams - Purdue

Austin Wiley - Auburn

Scottie James - Liberty

Tyrique Jones - Xavier

Cletrell Pope - Bethune-Cookman

Patrick Harding - Bryant

Chris Harris - Houston

Freddie Gillespie - Baylor

Myles Johnson - Rutgers

Loudon Love - Wright State

Austin Phyfe - Northern Iowa

Mubarak Muhammed - Louisiana Tech

Omer Yurtseven - Georgetown

Mahamadou Diawara - Stetson

Udoka Azubuike - Kansas

Josh Mballa - Buffalo

Brison Gresham - Houston

Nate Watson - Providence

RJ Williams - Boise State

Dave Bell - Jacksonville

Emanuel Littles - North Alabama

Nico Carvacho - Colorado State

James Dickey - UNC Greensboro

Aaron Carver - Old Dominion

Dwight Wilson - James Madison

James Butler - Drexel

John Mooney - Notre Dame

Nathan Knight - William & Mary

Joel Soriano - Fordham

Kai Edwards - Northern Colorado

Ahsan Asadullah - Lipscomb

Yoeli Childs - BYU

Elyjah Goss - IUPUI

Tyler Bey - Colorado

Carlos Dotson - Western Carolina

Lowest ORB%, Highest DRB%

    Since we're looking at some of the most elite rebounders in the country, no players are found in the bottom left corner as they're at least specialists in one of ORB% or DRB%, if not both. We can just skip that empty area and move up to the top left corner which is where defensive rebounding specialists are found. The two players we'll focus on from here are Tyler Bey and Yoeli Childs.

Tyler Bey - Colorado

    We've actually already looked at Bey in the article about defensive rebounding percentage vs. defensive rating. We saw that on top of being a great defensive rebounder, he was very good in other defensive attributes such as blocks (1.2 per game), steals (1.5 per game), and limiting fouls (1.9 per game). It makes sense why he has a lot of NBA draft intrigue due to his defense. According to the chart, his offensive rebounding percentage could be seen as a weakness as it's just over 9%. However, Colorado was only a slightly above average team in terms of ORB%, so offensive rebounding likely wasn't a huge priority for Bey. This is especially true given his 55.5% EFG% as his value as a shooter made it more costly to give up a potential open look from a kickout to try to get the rebound himself. 

Yoeli Childs - BYU

    Childs only managed to suit up for 19 games this past season, a byproduct of a 9-game suspension for an NCAA paperwork issue coupled with a dislocated finger. However, he was a tremendous asset to the Cougars when he did play, and was a big part in the team being seen as comfortably in the NCAA Tournament field and as having the potential to make a run. Since he missed so many games, Childs still ended up with an 8.8% ORB% despite having only gotten 38 total offensive rebounds during the season. This is well behind even Bey's low total of 62. Nevertheless, he still had the best ORB% on his team. Ultimately, this didn't matter too much because BYU shot the ball so ridiculously well that offensive rebounds weren't as critical for the team. Thus, they actually ended up 343rd in the nation in total offensive rebounds. While not super active on the offensive glass, Childs was relied upon to get defensive rebounds partially because of the lack of height on this BYU squad. The Cougars were actually quite a good defensive rebounding team aside from Childs, but him being the second tallest of players who played significant minutes at 6'8" provided a helpful boost.

Highest ORB%, Highest DRB%

    As we head over to the top right corner of the graph, we see some of the best out of this already elite group of rebounders. They truly excelled on both the offensive and defensive glass and dominated with their positioning and physicality. We'll focus on three players who are most clearly in this corner: Kevin Marfo and Patrick Harding (as promised in my last article) as well as Austin Wiley.

Kevin Marfo - Quinnipiac (now Texas A&M)

    Marfo topped the list of highest DRB% at 36.9% and wasn't too far off the top in terms of ORB% with his 19.9%. Given his dominance in both of these metrics, it's no surprise that he entered the abrupt beginning of the offseason as a highly-sought grad transfer. He ultimately decided to take his talent to College Station to play for Coach Buzz Williams and the Aggies. The big question is, will he be able to maintain such high rebounding percentages? The reason for optimism is that Marfo's incredible rebounding numbers didn't occur because he dominated in limited minutes. Rather, he was a 28.1 MPG game guy and grabbed 134 total offensive rebounds and 265 total defensive rebounds. It has to be mentioned that while he got both a large percentage and volume of rebounds when he was in the game, he was doing so in the MAAC. There's a clear step up in size and physicality of bigs going to the SEC that Marfo will now have to battle against to repeat such an incredible season. The only Power 5 competition that Marfo faced this past season was ACC foe Miami, and he had no trouble in getting 3 offensive boards and 10 rebounds on the defensive end. Unfortunately, Miami was a pretty bad rebounding team and so this doesn't really give us great insight as to whether Marfo can truly hold his own against higher level opposition. Overall, there's definitely evidence to support Marfo fitting right in and helping the Aggies improve upon their 30.9 rebounds per game this past season, but we'll have to wait until the first few games to truly know.

Patrick Harding - Bryant

    Harding achieved his awesome DRB% and ORB% through only 18.8 MPG, so almost 10 MPG less than Marfo. This past season, his sophomore year, Harding had a bit of a breakthrough in terms of rebounding. While his minutes only experienced a slight uptick (17.5 to 18.8 per game), he almost doubled his total rebounds and had 94 total offensive rebounds and 186 total defensive rebounds. Accordingly, his DRB% and ORB% rose to get him towards the top of both of these lists nationally. Judging from these percentages only, it seems as though Harding's performance on the boards has warranted a more dramatic minute increase than the one that he had this past season. One concern is that he only scored 4.7 points per game, but he did so with a very good 61.3% EFG%. Therefore, he should be able to fill more of a scoring role with his efficiency. Coach Jared Grasso's group is projected to improve upon their 7-11 conference performance from this past season, and I can see Harding playing a strong part in that.

Austin Wiley - Auburn

    Wiley had a pretty solid freshman season for the Tigers, but his college career took a strange turn when he was forced to sit out of the 2017-18 season due to a potential recruiting violation related to assistant Chuck Person. He couldn't have asked for a better season to return to as 2018-19 was when Auburn made its legendary Final Four run. He didn't have huge role on that team being only a 6.9 ppg and 4.0 rpg guy, but he was doing so on 13 minutes per night. This past season, he was given a larger role with 21.4 minutes per game and has certainly delivered. He got his scoring up to 10 ppg with an effective 57.4% EFG%. His biggest strength though, of course, was rebounding. His DRB% and ORB% both increased significantly and he's now one of the best rebounders in the nation according to this metric. Wiley certainly has bounced back from the challenges of 2017-18 to become an elite rebounder and member of a Final Four squad.

Highest ORB%, Lowest DRB%

    We finish off the analysis in the bottom right corner to look at offensive rebounding specialists. Like the defensive rebounding specialists above, just because these guys specialize more in one rebounding area, it doesn't mean that they're a liability on the glass on the other end. We'll look at both Chris Harris and Jordan Minor.

Chris Harris - Houston

    I mentioned Harris in my previous article about ORB% vs. offensive ratings when talking about another Houston player that can be found on the above graph, Brison Gresham. I mentioned how Harris has graduated and how his loss will be tough for the Cougars to replace because of his rebounding ability. Rebounding was a huge priority for last season's Cougars as they were part of a three-way tie for the most rebounds per game with 38.6. They also had the second best ORB% at 39%. Harris was a strong part of this as he had the best ORB% on the team. He almost doubled his total offensive rebounds from 42 in the 2018-19 season to 80 this past season. This worked out to be 2.7 per game on only 17.9 mpg. As mentioned in my previous article, guard Nate Hinton who's more of a defensive rebounding specialist is also leaving the program to go pro. The Cougars certainly have their work cut out for them if they want to reign atop the NCAA rebounding rankings once more.

Jordan Minor - Merrimack

    Minor was the ultimate newbie this past season having been a freshman on a Merrimack squad in its first season of Division I hoops. While not a huge scoring threat yet with only 6.2 points per game, Minor's 20.5% ORB% was a huge benefit to his team. This is because the Warriors weren't a great offensive rebounding team having only gotten 6.4 per game which contributed to a 20.4% ORB%. Somewhat surprisingly, Minor's individual ORB% was better than his overall team's which hasn't been the case for the aforementioned players. To further emphasize just how key Minor was to Merrimack's offensive rebounding, he himself got 2.6 of the 6.4 team offensive rebounds per game. It'll be interesting to see if Minor can continue to shoulder such a large portion of the offensive rebounding load, especially if he sees more time on the court. I think there's certainly a case to be made for this to happen. He had a 105.8 ORtg as a freshman which is pretty solid, and he showed flashes of scoring potential such as going 6-9 for 15 points against Jacksonville. The key for him will to be to maintain his excellent offensive rebounding ability while improving his game in other areas.

Overall Thoughts:

    I really enjoyed writing this three-part series of articles on DRB% and ORB% because it of course highlighted well-known elite rebounders such as Marfo who play a lot of minutes and thus have a large volume of rebounds. However, it's also nice to see credit given to players who simply make the most out of the more limited minutes that they're given, a player like Patrick Harding or Chris Harris. We saw that a lot of players in the graph are pretty good at both offensive and defensive rebounding which makes sense as the two skills are very similar. Yet there are a few clear specialists in either DRB% or ORB%, and we had to dive deeper to see why that was so. For instance, Childs was more of a defensive rebounding specialist because his team shot so well and offensive rebounds weren't as essential, whereas Chris Harris had many good defensive rebounding teammates like Nate Hinton and so he specialized more on the offensive end. Despite not being as flashy as scoring, rebounding is critical to the game of basketball, and all of these players deserve praise for excelling in this area.

Thanks so much for reading this far, I hope you enjoyed it! Be sure to check out some of my other articles!

2019-2020 Defensive Rebounding Leaders Analysis

    Welcome back! In my last article, I looked at the top 20 players in terms of offensive rebounding percentage (the percentage of available offensive rebounds a player got while he was in the game) and compared their ORB% to their offensive ratings (the number of points a player produces for every 100 possessions). If you missed the article, you can find it here. We saw that all of these players generally had very high O Ratings which were helped by their offensive rebounding. Yet, those who had the highest O Ratings generally were very good in other areas of offense as well such as shooting efficiency measured by effective field goal percentage (a modified field goal percentage where three-pointers count for more as they're worth more points).

    I felt like the next step would be to switch over to looking at defensive rebounding percentage (calculated the exact opposite of offensive rebounding percentage). You can probably already guess that I'll be comparing this to defensive ratings to see the role that defensive rebounding plays in DRtg. Defensive ratings are like offensive ratings except they're the points allowed by a player per 100 possessions. Like offensive ratings, there's a complicated formula behind them that I won't mention here. Also, they're less accurate than O Ratings because they're much more heavily influenced by how good a player's entire team is defensively. Still, it'll be interesting to look at. Like the last article, defensive rebounding percentage (DRB%) and D Rating data are from Fox Sports.

 Here's the graph comparing DRB% to D Rating:

Here's the complete list of players:

Kevin Marfo - Quinnipiac

RJ Williams - Boise State

Patrick Harding - Bryant

Dave Bell - Jacksonville

Emanuel Littles - North Alabama

Nico Carvacho - Colorado State

James Dickey - UNC Greensboro

Aaron Carver - Old Dominion

Dwight Wilson - James Madison

Austin Wiley - Auburn

James Butler - Drexel

John Mooney - Notre Dame

Nathan Knight - William & Mary

Joel Soriano - Fordham

Kai Edwards - Northern Colorado

Ahsan Asadullah - Lipscomb

Yoeli Childs - BYU

Elyjah Goss - IUPUI

Tyler Bey - Colorado

Carlos Dotson - Western Carolina

Lowest DRB%, Low DRtg

    To start off the analysis, we'll begin in the bottom left corner which is where we find the players with the lowest defensive rebounding percentage and the lowest defensive ratings. We'll focus on Tyler Bey and Joel Soriano. Even though these players have some of the lowest defensive rebounding percentages of the group, they're still only about 7% lower than the national leader, Kevin Marfo. Not too shabby. Since the lower the defensive rating is, the better (you want to allow as few points per 100 possessions as possible), let's see if other factors besides strong defensive rebounding has led to these players being such good defenders.

Tyler Bey - Colorado

    Bey is no stranger to the list of top defensive rebounders as he was second in the country last year in DRB% at 35.9%. His numbers came down a bit this season at 29.9%, but his overall defensive rating still improved. This is because Bey was able to double his steals per game from 0.8 to 1.6 and he maintained a steady 1.2 blocks per game. Another factor in the improved DRtg was that he became more disciplined in his fouling as he only averaged 2.6 fouls per game this season. Perhaps the most important factor though, was that Colorado as team got better defensively. As mentioned in the intro, a player's defensive rating is closely tied to his overall team's defensive rating. Colorado improved from allowing 97.5 points per possession in 2018-19 to allowing 94.6. Bey was well below the team average in both of these seasons, but the team improvement certainly helped him out a bit. It's clear that Bey has a strong defensive skill set beyond just rebounding as his steals, blocks, and foul discipline shows. It's no surprise that he took home the Pac-12 Defensive Player of the Year award and projects to be drafted in large part due to his defensive ability.

Joel Soriano - Fordham

    Unlike the established junior Bey, Soriano came off the bench and had fluctuating playing time for much of his freshman season for the Rams. However, he began to come into his own and started the last 11 games of the season. He only averaged 4.6 defensive boards per game, but in the limited minutes that he played this season, this translated to an excellent defensive rebounding percentage. Similarly, his block percentage was very solid (6.4%) despite him not having very many total blocks (only 0.9 per game). Offensively, Soriano struggled a bit with a 45.5% effective field goal percentage, but his efficiency improved as he experienced a larger role over those last 11 games. Soriano's excellent defensive numbers this season were largely driven by his DRB% and block% being so high in the restricted role that he had for most of the season. If he can maintain those rates and continue his better shooting efficiency while taking on more minutes this coming season, it will bode well for the team from the Bronx.

Lowest DRB%, Highest DRtg

    Now we move to the top left of the graph which is where players who had a tough time defensively, but still managed to excel on the boards appear. This would suggest that players in this region lean on defensive rebounding to make up for struggles in other defensive areas. The only player in this area to look at is IUPUI's Elyjah Goss, and we'll see what caused his DRtg to be so high.

Elyjah Goss - IUPUI

    At first glance, it's confusing to see how Goss could have such a high defensive rating. Nothing stands out as a glaring defensive weakness. He averaged 0.5 steals per game, 1.0 blocks per game, and 2.6 fouls per game. While these numbers don't necessarily wow you, they're not too far off from some of the numbers that our defensive standouts above were achieving. Thus, it's not the stats themselves that hurt Goss. Instead, it was partially the fact that he played 31.1 minutes per game for his team. His per-game stats ended up being very similar to Soriano above, but it took Soriano far less game-time to achieve them which is why he appears so much better from a DRtg standpoint. The more important factor though is that IUPUI wasn't a great defensive team last year, and we've already seen the influence of the team DRtg on an individual's rating. The Jaguars had one of the worst team defensive ratings at 111.7 points allowed per 100 possessions. Goss was actually well below this at 106.9, so while it was his worst DRtg in his three-year career thus far, we see that it certainly wasn't all his fault.

 Highest DRB%, Highest DRtg

    We're going to skip the top right corner as nobody falls within that region and we'll end this article with the bottom right. These are players who are clearly some of the best defensive rebounders in the minutes that they play and who also play solid overall defense to have such low defensive ratings. Four players could reasonably be looked at here (Bell, Harding, RJ Williams, and Marfo). I'm going to limit to the analysis today to just Bell and RJ Williams. Fear not Bryant and Texas A&M/Quinnipiac (if they don't hate him for leaving) fans, Harding and Marfo will be mentioned in a future article.

Dave Bell - Jacksonville

    Bell started his career in his home state of Ohio playing at Ohio State. After his sophomore season, he decided to take his talents down south to Florida to play for the Dolphins. He had a solid first season in the Atlantic Sun, but enjoyed a true breakthrough this past season. He experienced over a 16 point increase in his offensive rating (100.3 to 116.6) and over 9 point decrease in his defensive rating (98.4 to 89.0). As a coach, there's not much more you can ask for from your senior center. Defensively, he posted a very impressive 8.8% block rate, good for 43rd in the country. He also had 7.3 rebounds per game which is made more impressive by his high DRB%. It's a shame that Bell couldn't play another season to continue his upward trajectory. He'll certainly be a player that the Dolphins will miss next season.

RJ Williams - Boise State

    Williams came in at #2 of the players with the best defensive rebounding percentage  at 35.9%. Like Bell, he also transferred in for his final two seasons, but came from the JUCO ranks instead of another Division I program. He too experienced quite an impressive leap in both offensive and defensive ratings. His offensive rating climbed 18.6 (!) points and his defensive rating dropped by about 4 points to a nice 92.4 points allowed per 100 possessions. He actually experienced a slight decrease in both steals and blocks per game and even had an uptick in fouls per game. Thus, the improvement in DRtg seems to be driven a lot by Boise's team improvement from 2018-19's 103.3 points allowed per 100 possessions to this past season's 98.6. Williams still remained an above-average defender for his team this season and while not the focus of this article, his efficiency on offense was a tremendous help to the Broncos.

Overall Thoughts:

    Clearly, we see that individual DRtg depends far more heavily on team defensive success than ORtg relies on team offensive success. This dependency made analysis a bit more tricky as it was tough to tell how much a player's DRtg improvement was his own doing or was due to the team. Nevertheless, we got a glimpse of why Tyler Bey is becoming a frequently talked about draft prospect. We also saw a player in Soriano who looks to take on a larger role next season, and a player in Goss who will hope his team is able to get more on his level defensively. Lastly, we were able to identify two players in Dave Bell and RJ Williams who deserve recognition for really elevating their games in their final season of college hoops. 

Thanks so much for reading this far, I hope you enjoyed it!

2019-2020 Offensive Rebounding Leaders Analysis

    Welcome back! After spending the last article looking at skill curves and comparing usage rate to offensive rating for a few players, I decided to stick on the subject of offensive rating but with a different approach. I'll be looking players' cumulative offensive ratings for the entire season rather than on a per-game basis. As alluded to in the title, offensive rebounding will be the focus of this article, and so I'll be looking at the cumulative offensive ratings for the top 20 players in terms of offensive rebounding percentage. I decided to do this in order to see the role that offensive rebounding plays in a player's offensive rating.

    I chose to use offensive rebounding percentage instead of just each player's total offensive rebounds from the season because it adjusts for the fact that not every player has the same opportunity to get offensive boards. You'll notice that many players on this list didn't play starter's minutes, nor did they rack up a ton of total offensive rebounds. However, they took advantage of the time that they were given and dominated on the offensive glass when given the chance. I got both the offensive rebounding percentage and offensive rating data from Fox Sports. 

The following graph includes the top 22 (there was a three-way tie at 20) players in terms of offensive rebounding percentage and compares their offensive rebounding percentage to their season-long offensive rating:

As always, here's the full list of players and their schools:

Oscar Tshiebwe - West Virginia                   

Jordan Minor - Merrimack

Ed Croswell- La Salle

Kevin Marfo - Quinnipiac 

Trevion Williams - Purdue

Austin Wiley - Auburn

Scottie James - Liberty

Tyrique Jones - Xavier

Cletrell Pope - Bethune-Cookman

Patrick Harding - Bryant

Chris Harris - Houston

Freddie Gillespie - Baylor

Myles Johnson - Rutgers

Loudon Love - Wright State

Austin Phyfe - Northern Iowa

Mubarak Muhammed - Louisiana Tech

Omer Yurtseven - Georgetown

Mahamadou Diawara - Stetson

Udoka Azubuike - Kansas

Josh Mballa - Buffalo

Brison Gresham - Houston

Nate Watson - Providence

    At first glance, the graph seems very scattered and doesn't seem to tell us very much. After all, there's no clear trend such as the players with the highest offensive rebounding percentage also having the highest offensive ratings. However, if we look at the extremes (i.e. the corners), we can extract some useful information.

    For instance, the bottom right corner includes the players with some of the highest offensive rebounding percentages in this elite group, while also having some of the lowest offensive ratings. This suggests that these are players who struggle a bit in other areas of their offensive game, and their superb offensive rebounding abilities are inflating their O Rating a bit. For instance, Jordan Minor of Merrimack falls within this region. Minor, a freshman, wasn't a big part of the Warriors' strong first season in Division I having only played around 16 minutes per game off the bench. He only averaged 6.2 points and 4.4 boards per game, but his prowess on the offensive glass in the minutes that he was given helped drive his O Rating. 

    

    La Salle's Croswell also finds himself in this corner. The sophomore averaged more points (10.0) and shot a little better (60.2% eFG%) than Minor, albeit with more experience and minutes. With excellent offensive rebounding serving as a foundation, Croswell's shooting efficiency indicates that he might warrant an expanded offensive role. However, it won't be up to La Salle to provide such role anymore, as Croswell is heading to the Big East where he'll sit out a year and then suit up for Providence. This could be a really good get for the Friars.

    There's not really any players distinctly in the top right, so we'll move on to the bottom left. Of course, this is the area on the graph that you normally wouldn't want to be as it indicates the lowest offensive rebounding percentage and the lowest offensive rating. However, all of these players are the cream of the crop in offensive rebounding, and even the players in this corner have barely below a 105 O Rating which is still very solid. The two players that we'll focus on from this not so unfortunate after all section are Stetson's Mahamadou Diawara and Wright State's Loudon Love. 

    Diawara, a three star recruit, chose Stetson over offers from UMass, Rhode Island, Saint Joe's and Penn State per 247Sports. The choice seems to have paid off as he started the vast majority of the Hatters' games as a freshman and already has found himself on this elite list of offensive rebounders. The aspects that appear to be bringing down his O Rating are turnovers (3.0 per game), fouls (3.2 per game), and assists (0.7 per game). The influence of the turnovers and fouls is magnified a bit by the fact that he only played 26 minutes per game, but they should go down as he gains more experience. Getting assists up will be a little more tricky, but hopefully the coaching staff works with Diawara in getting him more involved with this aspect of the offense.

    Love's inclusion in this area surprised me a bit because he was the Horizon League Player of the Year this season. However, his O Rating was a little on the low side (at least in comparison to the rest of this group) partially because Love had his lowest eFG% (48.8%) in his career thus far. This can be explained by the fact that his usage rate was also the highest it's ever been at 31.9% as he was relied upon a ton to get the Raiders to a first place conference finish. His slight decline in shooting efficiency was countered by a nice jump in offensive rebounding percentage to help solidify himself on this list. Look for Love to be back on this list next year as he tries to end his senior season by doing the double again of  being Player of the Year and winning the league.

    Lastly, we move up to the top right. Austin Phyfe and Brison Gresham are the standouts here. This area suggests that these are players who are excellent offensive rebounders, but who also have many other offensive skills that have supported such high offensive ratings. 

    Phyfe had himself a tremendous season for Northern Iowa this year after coming back from a redshirt season due to a medical issue. He successfully made the transition from efficient sporadic starter in his freshman season to being a big piece on a great mid-major team this season. His offensive rating was 129.9 this season, good for 16th in the nation. This incredible O Rating was helped by an awesome 69.4% eFG% and Phyfe looks to be heading towards two more exciting seasons with the Panthers.

    Unlike Phyfe, Houston's Gresham didn't play a huge role on his team and only averaged 2.9 points per game. Yet the former UMass transfer made a key impact in his 15 minutes per game mostly off the bench with his offensive rebounding and blocking ability. Gresham's 1.2 blocks per game translates to 3.2 blocks per 40 minutes, which would've put him among the blocking elites this season. Aside from blocking on the defensive end, his 64.3% eFG% helped him also have such a high O Rating despite limited scoring. Teammate Chris Harris (also on this list of top offensive rebounders) is graduating and Nate Hinton, a guard lauded for his rebounding ability, is going pro which leaves the Cougars needing someone to step up. I think Gresham could be that guy.

    Overall, this graph showcased a lot of guys who I think have potential to be given a greater offensive role or who will continue to shine. I thought it was interesting too that each corner had its distinct identity of offensive player. As for my original question about what the impact of offensive rebounding percentage on O Rating is, I think its pretty clear that offensive rebounding percentage can be a big factor as all of these players excelled in this skill and none had an O Rating below 103. We saw with players like Phyfe and Gresham that a high offensive rebounding percentage provides a sort of minimum O Rating that then can be brought up to truly impressive levels with efficient shooting.

    

College Basketball Skill Curves

    Welcome back! Today I decided to look at skill curves for various college basketball players. What exactly is a skill curve you might ask? Unfortunately, I can't take credit for their creation as Dean Oliver discusses them in his 2004 book Basketball on Paper. The basic concept is to compare a player's offensive performance relative to how much is demanded of him when he's on the floor. This is achieved by comparing a player's offensive rating (a complicated formula that estimates how many points a player contributes for every 100 possessions) to his usage rate (the percentage of plays that a player is involved in). In his book, Oliver creates these graphs for NBA players only, so I decided to try to extend the idea to the college game.

     For example, here's what this type of graph looks like for Marquette's Markus Howard on a per game basis. This graph as with all following graphs is fitted with a trend line with varying levels of accuracy. In Howard's graph we see the data points are quite sporadic, but the trend line is still decent. We'll see later examples where this isn't the case. Also, in Oliver's book, he determines a year-by-year average O Rating for the NBA. In the absence of a predetermined average for college basketball this past season and the current limitations with how I'm getting my data, I think a rough estimate is probably somewhere around 105.

    As you can see, as Howard's usage rate increases, his offensive rating stays relatively the same. This is pretty unexpected as you would assume that as the player is tasked with doing more for the team, his offensive performance would start to decline due to fatigue, shooting inefficiency increasing with the volume, etc. While Howard's graph doesn't show the expected decrease, it doesn't mean that we can just assume that increasing Howard's usage rate up to say 70-80% will yield the same production. However, it does confirm what many already believe to be true: he can be a very effective high-volume offensive player. 

    As mentioned earlier, this is a case where the trend line comes out looking pretty strange, and suggests that Toppin's offensive rating steadily declines until his usage rate hits somewhere around 28%, and then suddenly his offensive rating gets briefly better. This isn't necessarily true, but the trend line does show the expected decline unlike Howard's above. This suggests that coach Anthony Grant was wise not to let Toppin's usage rate creep too much above 35% to preserve Toppin's incredibly productive and efficient season. This is emphasized by the fact that Toppin only had four games with an O Rating under the estimated average of 105 and two of those games were barely under this mark.

    After looking at two star players for their respective teams, I thought that it would be good to look at a more role-player type starter. Davide Moretti came to mind because I wanted to see the impact of his decision to leave Texas Tech to go pro. He certainly was valuable to the team as he had an O Rating above 105 in over half of his games. However, we see that his optimal usage rate is less than 25%, and so it shouldn't be too difficult for Tech to bring in a replacement with similar production at this less-demanding usage rate. It's certainly easier than a team like Marquette trying to replace Markus Howard whose optimal usage rate was much higher. Another reason for Tech fans to not be too upset is the fact that we see a downward trend line. It doesn't appear that Moretti was in for a breakout senior season offensively because his offensive performance declined as he was relied upon more.

    Unlike Moretti who doesn't look like he would've seen an offensive boost this coming season, we move into bench players and see that Alex O'Connell seemingly could be ready to make such a jump. Ignoring the two huge outliers on the right, O'Connell's O Rating generally improves as his usage rate goes up and peaks around 25% which is backwards. Like Toppin and Moretti (Howard is a weird exception), we should see O'Connell's O Rating steadily decline as his usage rate goes up, but that's just not the case. My guess is that it's because whenever O'Connell's usage for a game was over 20%, he likely played a lot of garbage-time minutes where he was actually able to do things on offense and help his O Rating. Regardless of reason, I'm curious to see how he would fare consistently having a usage rate of around 30%. Unfortunately, I don't see this happening next season even in spite of his transfer to Creighton given how strong the Blue Jays project to be. The only way we might get the chance to see if this skill curve is a fluke or not is if O'Connell can't get a waiver, and after sitting out a year sees a larger role awaiting him.

    Overall, these skill curves aren't the most informative because we're limited to the roughly 30 games that each player competed in this season. Players who missed games are even more difficult to assess with these graphs. Compare this to the NBA's 82 games and skill curves clearly paint a better picture for NBA data. Another downside is that Basketball on Paper is from 2004, leaving O Rating to be a bit dated. However, I think that skill curves still have some use for college basketball in that they told us what is generally assumed such as Markus Howard being a strong offensive asset despite being relied upon so much, and also some surprising information like Alex O'Connell potentially ready to see a vastly expanded role. I think that these skill curves could be particularly valuable in finding players ready for a breakout, as well as those who should see less usage. I know that the players that I selected for this article didn't necessarily show these trends (except possibly O'Connell), so if readers have suggestions for who to analyze next, my Twitter is @cbb_statistics. 

    Thanks so much for reading if you made it this far! If you enjoyed this article, be sure to check out some of my other ones such as my one on left versus right side layup percentages!

Left vs. Right Layup Percentages

      In my last article, I analyzed shooting volume and efficiency in the final 2 minutes for ACC players. I included all shot types for that analysis (except free throws), so three-pointers, two-point jumpers, layups, and dunks were all factored in. Today I thought it would be interesting to narrow down the shot type scope to just focus on layups. I started off by comparing the number of layup attempts from the left versus the right side to see if certain players dramatically favored a certain side. 

     I initially intended to do so with the top 20 Power 5 conference (plus Big East) guards with the most made layups. I wanted to focus specifically on guards because a guard layup generally involves more driving to the basket, and a big man layup centers more around post moves. I didn't want to have to compare these different types. Unfortunately, ESPN's shot tracking data was only deployed for certain games, and I needed this data to distinguish between layups from the left side of the basket versus the right side. There wasn't enough data for mamy of the top 20 guards with the most makes, so I resorted to looking at the top 20 players in terms of made layups with associated shot position data. This meant that I had to include both big men and guards in my top 20. Furthermore, it's evident that the shot tracking data was heavily used for games involving blue bloods, so you'll notice a lot of Duke, Kansas, Kentucky, etc. in the following list. While not what I was looking for at first, there's still some interesting insight to be drawn from this data.

Here's a more readable list of the the players with their respective teams also included:

Ashton Hagans - Kentucky

Collin Gillespie - Villanova

Davion Mitchell - Baylor

Devon Dotson - Kansas

Garrison Brooks - North Carolina

Jared Butler - Baylor

Jeremiah Robinson-Earl - Villanova

Jermaine Haley - West Virginia

Keyontae Johnson - Florida

Luka Garza - Iowa

Marcus Garrett - Kansas

Nick Richards - Kentucky

Payton Pritchard - Oregon

Tre Jones - Duke

Tyrese Maxey - Kentucky

Udoka Azubuike - Kansas

Vernon Carey Jr. - Duke

Will Richardson - Oregon

Xavier Tillman - Michigan State

Zavier Simpson - Michigan

   To reemphasize, this graph doesn't show who had the most layup attempts in the P5+BE this season, rather it shows the layup leaders for shots that actually had positioning data. Furthermore, these aren't even all of each player's layups because shot tracking data wasn't available for all of their games. Nevertheless, there's large enough samples to make at least some analysis.

    Devon Dotson, Marcus Garrett, and Vernon Carey Jr. stand out as the combined L/R volume leaders. Garrett's attempts were almost perfectly balanced at 63 from the left and 65 from the right. Dotson's 166 total layup attempts favored the right side slightly more, and we see that Vernon Carey Jr. clearly preferred the right with 79 attempts versus 59 from the left. Carey Jr. definitely had the biggest disparity between left and right, as most players were pretty balanced. We see this balance continued in the fact that there was a relatively even amount of players who preferred one side over the other. Overall, there wasn't as much disparity as I anticipated.

    Now that we discussed volume, it's time to dive deeper into just how efficient each player was on each side. First we'll start with the layup percentage from the left side:

     Florida forward Keyontae Johnson leads the way from the left having shot a blistering 28/32 (87.50%). We see him joined by fellow big men like Nick Richards (70.83%) and Xavier Tillman (71.88%) who also shot well from the left. From the guards, Collin Gillespie (65.00%) and Tyrese Maxey (70.59%) stand out. Maxey's Kentucky teammate, Ashton Hagans, struggled at 34.15% along with Baylor's Davion Mitchell (40.00%). Overall, the 20 players averaged 60.04% from the left.

Now, moving on to the right:

    Udoka Azubuike was the most efficient from the right and shot just slightly worse (84.62%) than Keyontae Johnson's 87.50% from the left. Again, we see Xavier Tillman up there with 75.00%. There weren't any major standouts from an inefficiency standpoint with Baylor's Jared Butler shooting the worst at just under 50%. Overall, the players were slightly more accurate from the right with an average of 62.32%.

    While it was interesting to look at each side in isolation, it's more useful to compare the difference between the two sides. This tells us if a player is dramatically more efficient from a specific side.

    This graph was created by taking the left-side layup shooting percentage minus the right-side percentage. Thus, players with a negative difference shoot better from the right, and those with a positive difference shoot better from the left.

    Ashton Hagans stood out in the left-side layup graph regarding how low his shooting percentage was, and this graph emphasizes how much more accurate he was from the right. Hagans' defense is what will get him drafted, but this left vs. right layup inconsistency could be something that teams look into and will be something he'll want to work at minimizing. Azubuike's left-side weakness is less of a concern because it's more of a product of how ridiculously high his right-side percentage was. Furthermore, he was a dunking machine throughout his college career which helped him lead the nation in field goal percentage this past season. A slightly below average left-side percentage (53.57%) helps illuminate just how efficient he was shooting elsewhere. The next two most significant players in terms of being better from the right are Davion Mitchell and Garrison Brooks, both of whom will be returning to school and will look to improve from the left.

    Moving to players vastly more efficient from the left, Keyontae Johnson's apparent right-side weakness is driven by his extremely high left-side percentage of 87.50% and is essentially the opposite of Azubuike in that regard. Unlike Hagans and Azubuike who are hoping to be drafted in mid-October, Johnson will be returning to Florida in the fall. This means that this disparity could be something for opponents to pay attention to in trying to force him to go right, unless they're willing to test their luck with someone who shot over 85% from the left this past season (at least from the sample we have).

    The last notable player to discuss is Iowa's Luka Garza. While the focus previously has been on players who shot much better on one side versus the other, Garza did the opposite and shot exactly the same from both sides at 66.67%. He even did so with different volumes from each side with 42 attempts coming from the left and 48 from the right. Many Hawkeye fans were angry that Garza was beat out by Obi Toppin for most of the major awards this season. Yet, he may get a shot at redemption if he chooses to come back for his senior year and will look to repeat this well above average layup efficiency from both sides en route to another dominant season. Again, these are not his total season numbers so it's very likely he didn't actually shoot this equally for the whole season, but it's still an interesting quirk from the sample we have.

    Overall, this left versus right layup analysis could've been done in a much more complete and accurate way via studying film. Of course, this method would avoid the issue of not having complete data for any of the players. Also, it's very likely that some of these layups were recorded incorrectly (such as for the wrong side) with the shot tracker data, and this questionable accuracy would've been reduced by manually compiling the data. However, this obviously would take a very long time, and so this quick and slightly more inaccurate method works well enough. This analysis showed us a little cause for concern for players with a large efficiency disparity like Hagans (and slightly less so for Azubuike) as they begin their NBA journeys. It also highlighted players such as Davion Mitchell, Garrison Brooks, and Keyontae Johnson who could really benefit with extra practice from their weaker side to avoid their inconsistencies being further exposed this coming college hoops season.