Basketball Hour With Buzz and Homer - February 26

[wadesworld]

If that's the case, it should be noted that Buzz specifically mentions Derrick Wilson's outstanding on ball defense and his ability to guard his man.  That doesn't sound like a 'sum is greater than it's parts' logic.

I'm not trying to argue that Otule shouldn't see the court.  He is our second most effective post player on the team.  I'm just tired of the conventional wisdom ignoring concrete data in favor of the ever popular 'eye test' as we know how effective that is.

Here's the problem: The game isn't played on paper.  How are these numbers determined?  Does it take into account team defense?  I have my doubts, as that is the type of things that doesn't have a number attached to it.  If the numbers are what we should believe, I have a hard time that someone who makes $2,000,000/year doing this for a living would not be aware of this.  Maybe I'm crazy for thinking that, but...

[tower912]

Buzz clearly uses a defensive metric than you do, mattyv.   He has mentioned Davante's and Deonte's defense in a negative light multiple times.   Actually watching the games, watching the Oxtule combination when Davante is forced to guard a perimeter player, watching what happens when Burton is guarding the post and fails to get a steal, watching Gardner by the last guy back while pressing, shows that neither is an accomplished defender.    I appreciate your pursuit of statistics.    They aren't always borne out on the court.   And I don't think it is a lack of effort anymore, though I did earlier this year with Gardner.    He simply isn't quick or tall.   Deonte is learning, but isn't always in the right place and isn't tall enough or quick enough to compensate.    Yet.   

[jesmu84]

If that's the case, it should be noted that Buzz specifically mentions Derrick Wilson's outstanding on ball defense and his ability to guard his man.  That doesn't sound like a 'sum is greater than it's parts' logic.

I'm not trying to argue that Otule shouldn't see the court.  He is our second most effective post player on the team.  I'm just tired of the conventional wisdom ignoring concrete data in favor of the ever popular 'eye test' as we know how effective that is.

you'll notice I spoke about the 5 spot in terms of team defense, not the 1. Also, i wasn't arguing over data/eye test. I was expanding upon what buzz has said. If you want Buzz's evidence, email him.

[BenCat12]

The best defensive lineups aren't determined by a single statistic or an eye test.  It should be determined by the opponent.  Chris is the much better option, defensively, against teams that have good dribble drive guys because he is able to alter shots.  Davante is the better option when a team has a traditional back to the basket scoring center IMO.  Davante was much more effective against Stainbrook, from Xavier, than was Otule.  Otule is better against teams like St. Johns, because our guards have a hard time staying in front of their man and Obekpa isn't a great post scorer.  Both have their positives and negatives.  No one guy is better in all situations.  Based on the majority of opinions on this board, Davante is better than he is given credit for and Chris is worse.  Mattyv1908's numbers also verify this.  Again, Buzz has to find all the right pieces each night because this roster(which he put together) has players that are too one dimensional.

[mattyv1908]

Here's the problem: The game isn't played on paper.  How are these numbers determined?  Does it take into account team defense?  I have my doubts, as that is the type of things that doesn't have a number attached to it.  If the numbers are what we should believe, I have a hard time that someone who makes $2,000,000/year doing this for a living would not be aware of this.  Maybe I'm crazy for thinking that, but...

Good question.  I'll attempt to answer as honestly as I can.

Baseball was the first sport to really delve into objective numbers back in the late 80's with sabermetrics.  It is head and shoulders ahead of the other sports when it comes to using statistics that measure a player's impact over statistics that measure a player's result through traditional statistical categories (HR's, RBI's, BA, etc).  In fact the book (and subsequent movie) 'Moneyball' most famously showed it's impact on the Oakland A's.  The Red Sox have implemented building a team around it to the tune of 2 World Series titles.  Tampa Bay, Minnesota and a couple other small market teams have utilized similar statistics to consistently build contending teams on small budgets.  It has had the single biggest impact on the sport since the invention of lights allowing for night games.

Basketball has trailed Baseball in using these types of statistics, but there have been many advances in the advanced metrics to calculate a player's value to his team's W/L record as well as his team's points/possession which are two of the best indicators of evaluating individual players.

Offensive Rating -  .96 X (FG attempts - off rebounds + turnovers +(.44 X FT attempts))

This stat simply measures the points scored over 100 possessions that a player is on the court.  It takes pace of play and every other subjective number out of play.  It measures how efficient a player's team is at scoring when they are on the court over 100 possessions.

Defensive Rating - essentially the same stat except it measures your opponents' points scored over 100 possessions that a certain player is on the court.

These along with offensive/defensive win shares (which could take up a semester to fully grasp) are a much better indicator than traditional statistical measures that fill up stat sheets on what impact and value a player has for his team on both sides of the court.

As to why a college or NBA coach for that matter would not look to the data there can be several reasons.  Just like any advancements in society, people adjust at different rates.  In this particular case I think it's not a question of whether these stats are reliable so much as how much does a particular coach trust the data he has at his disposal.  I should note that the statistics are not 100% predictive, especially over a short sample size like a single game or single TV timeout.  Basketball and it's nature results in substitutions due to fatigue so it's necessary that players receive rest.  It's impossible to have your five most efficient players on the court at all times.

What the data shows from a statistical standpoint is that Marquette is allowing 65.1 possessions per game as a team.  Davante Gardner's defensive rating of 102.7 would yield 66.86 opposing points per game if he played every minute.  Chris Otule's defensive rating of 101.7 would yield 66.21 opposing points per game if he played every minute.  It's hardly a difference on what value they bring to their opponents' point total per game.  Marquette allows 67.1 points/game as a team so both players are better individually than the team is collectively this season.

I can't speculate on how much if any of the advanced statistics factor into Buzz's decision making or how much he values these numbers in relation to a multitude of other things that do include individual merit on a subjective measure based on how he runs his program and what he chooses to reward.  I'm simply pointing out what the data bears out.

The reason I bring these numbers into focus is because Davante Gardner is our most efficient offensive player by a large margin.  It's not close.  And not just compared to Otule, but compared to everyone on the team he is hand's down our best offensive player.  I simply wanted to point out to the board that this talk of him being a defensive liability is a myth when compared to Otule and yet his offensive productivity is absolutely needed.  The numbers would indicate he should be on the floor with the sole exceptions being to keep him fresh and avoiding fatigue and in late game situations when offensive and defensive substitutions can be made due to the amount of dead ball clock stoppages.

[wadesworld]

Good question.  I'll attempt to answer as honestly as I can.

Baseball was the first sport to really delve into objective numbers back in the late 80's with sabermetrics.  It is head and shoulders ahead of the other sports when it comes to using statistics that measure a player's impact over statistics that measure a player's result through traditional statistical categories (HR's, RBI's, BA, etc).  In fact the book (and subsequent movie) 'Moneyball' most famously showed it's impact on the Oakland A's.  The Red Sox have implemented building a team around it to the tune of 2 World Series titles.  Tampa Bay, Minnesota and a couple other small market teams have utilized similar statistics to consistently build contending teams on small budgets.  It has had the single biggest impact on the sport since the invention of lights allowing for night games.

Basketball has trailed Baseball in using these types of statistics, but there have been many advances in the advanced metrics to calculate a player's value to his team's W/L record as well as his team's points/possession which are two of the best indicators of evaluating individual players.

Offensive Rating -  .96 X (FG attempts - off rebounds + turnovers +(.44 X FT attempts))

This stat simply measures the points scored over 100 possessions that a player is on the court.  It takes pace of play and every other subjective number out of play.  It measures how efficient a player's team is at scoring when they are on the court over 100 possessions.

Defensive Rating - essentially the same stat except it measures your opponents' points scored over 100 possessions that a certain player is on the court.

These along with offensive/defensive win shares (which could take up a semester to fully grasp) are a much better indicator than traditional statistical measures that fill up stat sheets on what impact and value a player has for his team on both sides of the court.

As to why a college or NBA coach for that matter would not look to the data there can be several reasons.  Just like any advancements in society, people adjust at different rates.  In this particular case I think it's not a question of whether these stats are reliable so much as how much does a particular coach trust the data he has at his disposal.  I should note that the statistics are not 100% predictive, especially over a short sample size like a single game or single TV timeout.  Basketball and it's nature results in substitutions due to fatigue so it's necessary that players receive rest.  It's impossible to have your five most efficient players on the court at all times.

What the data shows from a statistical standpoint is that Marquette is allowing 65.1 possessions per game as a team.  Davante Gardner's defensive rating of 102.7 would yield 66.86 opposing points per game if he played every minute.  Chris Otule's defensive rating of 101.7 would yield 66.21 opposing points per game if he played every minute.  It's hardly a difference on what value they bring to their opponents' point total per game.  Marquette allows 67.1 points/game as a team so both players are better individually than the team is collectively this season.

I can't speculate on how much if any of the advanced statistics factor into Buzz's decision making or how much he values these numbers in relation to a multitude of other things that do include individual merit on a subjective measure based on how he runs his program and what he chooses to reward.  I'm simply pointing out what the data bears out.

Pretty interesting.  I have never gotten into advanced metrics in any sport, but just from the untrained eye it seems as though in basketball it is comparable to a player's +/- in that it doesn't really measure the effectiveness of one individual player, it measures the effectiveness of everyone on the court at the time one player is on the court.  I guess to me it just makes more sense that advanced metrics would be more valuable and accurate in a game like baseball where it is more often an individual vs. an individual matchup (batter vs. pitcher, etc.) and doesn't rely on other people on the court/field altering the output of the data.  Again, this is just from the untrained eye.  To me, it is harder to buy into it when there is so much dependent on 4 other guys on the floor in basketball.  I'm sure it definitely can help determine who is better in what situations, but there are too many outside factors contributing to those numbers.

[mattyv1908]

Here's Gardner and Otule's offensive ratings.

Gardner 127.3 / 100 possessions X 69.1 offensive possessions/game = 87.96 points/game
Otule 110.5 / 100 possessions X 69.1 offensive possessions/game = 76.36 points/game

So Gardner's defensive liability is roughly .5 points/game more than Otule but Otule's offensive liability is roughly 11.5 points/game less than Gardner's.

I'd love for someone to show me some empirical evidence that would indicate this team is better with Gardner sitting and Otule playing other than when necessary.  Using any form of subjective measures including our coach's decisions do not count as none of that is based in statistics.  Remember these numbers are not projections they are actually derived from this season's games played.  The sample size is great enough to take out any variables that would skew them.

[mattyv1908]

Pretty interesting.  I have never gotten into advanced metrics in any sport, but just from the untrained eye it seems as though in basketball it is comparable to a player's +/- in that it doesn't really measure the effectiveness of one individual player, it measures the effectiveness of everyone on the court at the time one player is on the court.  I guess to me it just makes more sense that advanced metrics would be more valuable and accurate in a game like baseball where it is more often an individual vs. an individual matchup (batter vs. pitcher, etc.) and doesn't rely on other people on the court/field altering the output of the data.  Again, this is just from the untrained eye.  To me, it is harder to buy into it when there is so much dependent on 4 other guys on the floor in basketball.  I'm sure it definitely can help determine who is better in what situations, but there are too many outside factors contributing to those numbers.

The reason myself and a few others here discount +/- stats is because those have been proven to be completely random from season to season when NBA all stars have almost identical stats from year to year and yet their +/- numbers swing wildly.

You are correct in that basketball does rely on other players, or more correctly, an individual's numbers are effected by the positive and negative things that occur around him as well.  Where they are extremely accurate is when you take an individual's offensive or defensive efficiency and calculate it out based on the team's average number of possessions and then reference it to the team's points allowed and points scored averages.  You start to see who is helping and hurting the team with their contributions compared to the team averages in a much more effective way than +/- ever could.

[Henry Sugar]

Here's Gardner and Otule's offensive ratings.

Gardner 127.3 / 100 possessions X 69.1 offensive possessions/game = 87.96 points/game
Otule 110.5 / 100 possessions X 69.1 offensive possessions/game = 76.36 points/game

So Gardner's defensive liability is roughly .5 points/game more than Otule but Otule's offensive liability is roughly 11.5 points/game less than Gardner's.

I'd love for someone to show me some empirical evidence that would indicate this team is better with Gardner sitting and Otule playing other than when necessary.  Using any form of subjective measures including our coach's decisions do not count as none of that is based in statistics.  Remember these numbers are not projections they are actually derived from this season's games played.  The sample size is great enough to take out any variables that would skew them.

I'm not going to argue the point about Gardner vs Otule. However, there are different stats showing defensive performance. There's the Dean Oliver approach, which also has limitations. And then there's what Buzz uses, which is the data from Synergy.

I've tried to get Synergy access, but couldn't. Synergy told me it was only for coaches. However, somehow the Paint Touches guys have it.

From last year
http://painttouches.com/2013/03/12/postseason-statistical-primer-marquette-defense/

From the first three games of this year
http://painttouches.com/2013/11/18/marquette-synergy-individual-defensive-numbers/

The numbers from last year are much better to look at. These types of numbers show why Buzz favors certain players and not others. There is a much larger disparity between defensive stats in Synergy.

[Galway Eagle]

Buzz is a sauce guy? This changes everything! PreBBQ or post, what he uses it as a glaze during?! So many questions!!!

[brandx]

I've tried to get Synergy access, but couldn't. Synergy told me it was only for coaches. However, somehow the Paint Touches guys have it.

I'm guessing if you want to pony up the bucks, you could get the info - but it's gonna be a real big pony. These services that provide info to teams charge pretty steeply.

[mattyv1908]

Buzz clearly uses a defensive metric than you do, mattyv.  Regardless of Buzz's evaluation criteria, it doesn't change the metric used by NBA executives or college coaches alike and while it doesn't mean Buzz is wrong, it doesn't make him correct by default either

He has mentioned Davante's and Deonte's defense in a negative light multiple times.  Again, this doesn't mean what you're implying it to mean either.  In his comments to Homer Buzz himself stated he's not a good coach.  Should we believe him and look for someone else just because Buzz said?

Actually watching the games, watching the Oxtule combination when Davante is forced to guard a perimeter player, watching what happens when Burton is guarding the post and fails to get a steal, watching Gardner by the last guy back while pressing, shows that neither is an accomplished defender.  I bolded the above because seeing isn't always believing.  It's been proven that human testimony is actually very inaccurate, as we have trouble remembering details and then the narrative we want to believe is blended into our recollection of the events.  You're post is a projection of your recollection of what you watched, not based in anything factual other than your memory.

I appreciate your pursuit of statistics.  Yet I'm throwing them out the window in favor of what I watched.

They aren't always borne out on the court.  Actually they're entirely founded in what happened on the basketball court.

  

[oldwarrior81]

does the defensive possession calculations actually use the number of defensive possessions a player was on the court during, or is it a by product of the game totals vs minutes played?

The way Buzz substitutes the bigs I would guess Otule may see more defensive possessions per minute played than Gardner does.

That substitution pattern is what essentially makes the +/- numbers for Gardner/Otule flawed.  

[NickelDimer]

Stats are for losers...

But those metrics are interesting

[brandx]

Stats are for losers

I assume that's sarcasm - but...

If you are talking about using stats to argue over something that happened, you are correct.

But, if you are using stats to prepare for the next game or season, you would be incorrect.

[mu03eng]

Stats are for losers...

But those metrics are interesting

Agreed, everyone knows 67.3% of statistics are made up on the spot.

[willie warrior]

Here's Gardner and Otule's offensive ratings.

Gardner 127.3 / 100 possessions X 69.1 offensive possessions/game = 87.96 points/game
Otule 110.5 / 100 possessions X 69.1 offensive possessions/game = 76.36 points/game

So Gardner's defensive liability is roughly .5 points/game more than Otule but Otule's offensive liability is roughly 11.5 points/game less than Gardner's.

I'd love for someone to show me some empirical evidence that would indicate this team is better with Gardner sitting and Otule playing other than when necessary.  Using any form of subjective measures including our coach's decisions do not count as none of that is based in statistics.  Remember these numbers are not projections they are actually derived from this season's games played.  The sample size is great enough to take out any variables that would skew them.

That's because nobody can. Almost all of those Otule supporters do so because of his "story" It is inspirational. Gardner is the "fat" kid that everybody likes to criticize. Gardner has improved himself more in 4 years than Otule has in 6. Otule is only slightly better defensively than Gardner, but many consider him way ahead, because of their bias. Otule gets the nod on defense because he blocks shots better-- and that is because he usually has 3 or more inches than anybody else on the floor.
Otule seals off more than Gardner for others but Gardner also does that. Gardner is a better rebounder, better shooter from mid range, better FT shooter, better passer, better dribbler, better scorer, and has better hands. They both help the team, but Gardner deserves more minutes and is definitely the guy to be in at crunch time.

[PGsHeroes32]

I like what Ben cat said.

Chris is definitely better at altering shots, taking charges and all around stoping a drive.

Ox is the far better 1 on 1 defender where the other team gives their guy and tells him to go to work. Davante is big and can't be pushed around.

It is true that some people here act like davante is some horrible defender when he's actually decent. Obviously Chris has his positives as well.

[77ncaachamps]

Meet Bo McCalebb: one of the best players Buzz has been around.

http://www.youtube.com/watch?v=3IBi9iwWrng

[BallBoy]

The best defensive lineups aren't determined by a single statistic or an eye test.  It should be determined by the opponent.  Chris is the much better option, defensively, against teams that have good dribble drive guys because he is able to alter shots.  Davante is the better option when a team has a traditional back to the basket scoring center IMO.  Davante was much more effective against Stainbrook, from Xavier, than was Otule.  Otule is better against teams like St. Johns, because our guards have a hard time staying in front of their man and Obekpa isn't a great post scorer.  Both have their positives and negatives.  No one guy is better in all situations.  Based on the majority of opinions on this board, Davante is better than he is given credit for and Chris is worse.  Mattyv1908's numbers also verify this.  Again, Buzz has to find all the right pieces each night because this roster(which he put together) has players that are too one dimensional.

I don't believe that either Otule or Gardner are world beaters when it comes to the defensive end of the court.  Neither can truly guard against perimeter threats.  I believe Otule is better defender overall due to athleticism, speed, and stamina.  Gardner does not play nearly as effective against athletic centers or in transition.  Folks like Stainbrook do not expose him because the two are similar players.  St John's and Creighton play against Gardner's strengths and therefore do expose him.

Buzz takes this into account for every game and switches playing time accordingly.  He also will look to do the offensive/defensive switch.  If you merged the two together we would have a very dominant center.