Marquette and Priorities - Part Two
Source: Marquette and Priorities - Part Two (http://www.crackedsidewalks.com/2014/03/marquette-and-priorities-part-two.html)
This is a Five Part series looking more in-depth at Marquette and the Priorities on the court.
- Part One - What does it take to make the Final Four?
- Part Two - Which of the Four Factors is more important than the other three combined? (spoiler: it's eFG%)
- Part Three - Marquette is not good enough on eFG% to have a top 20 offense/defense
- Part Four - What goes into eFG% for Top 20 teams?
- Part Five - Are Paint Touches the root cause?
In Part One of this series, we presented the idea that a team needs roughly a top 20 offense and a top 20 defense to make the Final Four.*
*Or have some sort of weird super-lucky runPart Two - Which of the Four Factors is more important than the other three combined? (spoiler: it's eFG%)Without boring everyone, it's possible to make a simple model that predicts efficiency. Here is the model.
(http://i228.photobucket.com/albums/ee41/roblowe14/Whatmakesefficiency.png) (http://s228.photobucket.com/user/roblowe14/media/Whatmakesefficiency.png.html)
For example, take the factor under question (eFG%) and then multiply by 1.79 for that component's contribution to efficiency. Repeat for the other factors and add them all together.
Here's how this model tracks for this season. (It actually compares really well for all seasons, but then the chart begins to look like an arrhythmia). Pretty close, right?
(http://i228.photobucket.com/albums/ee41/roblowe14/Whatmakesefficiency3.png) (http://s228.photobucket.com/user/roblowe14/media/Whatmakesefficiency3.png.html)
Once you adjust for the fact that turnovers are a negative contribution to efficiency, each aspect adds up to the following percentages.
(http://i228.photobucket.com/albums/ee41/roblowe14/Whatmakesefficiency2.png) (http://s228.photobucket.com/user/roblowe14/media/Whatmakesefficiency2.png.html)
That's right - eFG% is 63% of the total contribution to efficiency. eFG% is more important than all of the other three factors combined. This concept should be in ALL CAPS and bolded and underlined and have approximately 400 exclamation points!!!
(http://media.tumblr.com/8ac7995f3d066b680e0857da5560fd32/tumblr_inline_mua0pr5v501r5696t.gif)
This means that it is practically impossible to be a top 20 offensive or defensive team without also being good at eFG%. If you want the other numbers, Turnover Rate is 12%, Offensive Rebounding Percentage is 19%, and Free Throw Rate is 5%. Like those matter anymore.
For example, if a team like St. Johns averages an eFG% of 49% and Marquette allows them to shoot 56% at the Bradley Center, that's not so good.
SummaryeFG% is 63% of the total contribution to efficiency and it's more important than all of the other factors combined.
In Part Three, we will show that Marquette is not good enough consistently on eFG% to have a top 20 offense or defense.
Great stuff, really enjoying this whole series. I'm a little stuck on how you get from the regression coefficients to the %contribution to efficiency? I thought maybe you were just dividing the multiplier by the sums but that doesn't seem to work out.
In this model eFG% is a combination of both Offensive eFG% and Defensive eFG% correct? Do you just average them?
Quote from: THRILLHO on March 15, 2014, 06:16:29 AM
Great stuff, really enjoying this whole series. I'm a little stuck on how you get from the regression coefficients to the %contribution to efficiency? I thought maybe you were just dividing the multiplier by the sums but that doesn't seem to work out.
Multiply the coefficient by the factor value.
Use the absolute value because turnovers are negative
Sum all four products together.
Divide each product by the absolute sum
I will also send you the actual spreadsheet later
Quote from: mu03eng on March 15, 2014, 08:12:38 AM
In this model eFG% is a combination of both Offensive eFG% and Defensive eFG% correct? Do you just average them?
It's just offensive eFG%, but the defensive numbers are consistent
Sugar - So is it safe to assume based on your findings that putting a team on the floor who has a much better ability to put the ball in the basket, as opposed to being a better protector of the ball - is significantly more important?
Has Buzz gotten too obsessive over worrying about turnovers? I know he references "it's a possession game," and a turnover obviously results in one less possession - yet at what point does he not look at the data and say - there are an awful lot of empty possessions over the course of 40 minute game that are the result of playing 4 on 5 offensively?
What is the opportunity cost of playing a guy who will rarely take a 3 point shot as a PG, and if so, converts 7% of them - and when the opposition knows this - and sags off that player what is the trickle down effect on limiting efficiency elsewhere among the other guys?
Just doesn't seem like rocket science...
If Buzz is the numbers guy people claim he is one would believe he's looked at the same data. Perhaps he thought D Wil and J Wil would snap out of their year long funks.
Quote from: Ners on March 15, 2014, 11:04:28 AM
Sugar - So is it safe to assume based on your findings that putting a team on the floor who has a much better ability to put the ball in the basket, as opposed to being a better protector of the ball - is significantly more important?
Has Buzz gotten too obsessive over worrying about turnovers? I know he references "it's a possession game," and a turnover obviously results in one less possession - yet at what point does he not look at the data and say - there are an awful lot of empty possessions over the course of 40 minute game that are the result of playing 4 on 5 offensively?
What is the opportunity cost of playing a guy who will rarely take a 3 point shot as a PG, and if so, converts 7% of them - and when the opposition knows this - and sags off that player what is the trickle down effect on limiting efficiency elsewhere among the other guys?
Just doesn't seem like rocket science...
I'm not going to get directly into your leading question. However, I will say that MU's ranking for each factor is in Part Three of the series.
Quote from: muhoops1 on March 15, 2014, 02:45:04 PM
If Buzz is the numbers guy people claim he is one would believe he's looked at the same data. Perhaps he thought D Wil and J Wil would snap out of their year long funks.
I personally don't think he has looked at the same data.
Quote from: Ners on March 15, 2014, 11:04:28 AM
Sugar - So is it safe to assume based on your findings that putting a team on the floor who has a much better ability to put the ball in the basket, as opposed to being a better protector of the ball - is significantly more important?
Has Buzz gotten too obsessive over worrying about turnovers? I know he references "it's a possession game," and a turnover obviously results in one less possession - yet at what point does he not look at the data and say - there are an awful lot of empty possessions over the course of 40 minute game that are the result of playing 4 on 5 offensively?
What is the opportunity cost of playing a guy who will rarely take a 3 point shot as a PG, and if so, converts 7% of them - and when the opposition knows this - and sags off that player what is the trickle down effect on limiting efficiency elsewhere among the other guys?
Just doesn't seem like rocket science...
A turnover is not losing a possession. It is the result of a possession.
Theoretically a team that has 20 possessions turns it over 10 of them and makes 10 3 pointers is more efficient offensively than a team that has 20 possessions without any turnovers but shoots 8 for 20 in those 20 possessions.