Do you think there is any value in creating an adjusted turnover metric? For example, one that might incorporate free throw misses as turnovers, etc.
If you do think there is value, what sort of data would you incorporate into the metric?
If you don't think there is any value, do you think it is because most of the value of creating this adjusted metric absorbed in a statistic like points per 100 possessions? Or because how we already measure turnovers is sufficient?
Adjusted Turnover Metric?
Re: Adjusted Turnover Metric?
Joe Sill (now with the Wizards) was the first to produce adjusted plus minus in public at the factor level. I had expressed in that level of detail for several years before but was not up to attempting to produce it. Evan Zamir has done it more recently. http://www.d3coder.com/thecity/a4pm/
Re: Adjusted Turnover Metric?
Are 2 missed free throws different from a missed FG? Would you incorporate missed FGs too?
Re: Adjusted Turnover Metric?
I think what jaebradley is asking, is for a player's "0-points-possessions." As a function of player plays used, this is equal to (1-Floor%). Anyone got the formula for Floor%?
Re: Adjusted Turnover Metric?
Do you think there is any value in creating an adjusted turnover metric? For example, one that might incorporate free throw misses as turnovers, etc. If you do think there is value, what sort of data would you incorporate into the metric?
I think what jaebradley is asking, is for a player's "0-points-possessions."
what i refer to as a zero point possession (my definition may be different than others) is a team possession where a specific player on offense is personally responsible for his team not scoring, by either missing a FGA that is rebounded by the defense, by missing a FTA that is rebounded by the defense, or by committing a turnover. these can be estimated for a player (or counted using play-by-play)...
i then divide a player's points scored by his zero point possessions to get pts/0ptposs, which i use as a measure of offensive efficiency. so for example, this past season deandre jordan shot better overall than did kawhi leonard, a 59.1% vs. a 58.5% ScFG% (2s, 3s, and FTs). but jordan turned the ball over with 10% of his touches, leonard with just 5% of his touches. when you work out the numbers leonard had pts/0ptposs of 2.58, jordan 2.06. in other words, for each time he cost his team a team possession without scoring (and a team possession was worth on average about 1.05 pts this season), leonard scored 2.58 pts, jordan just 2.06 pts...
I think what jaebradley is asking, is for a player's "0-points-possessions."
what i refer to as a zero point possession (my definition may be different than others) is a team possession where a specific player on offense is personally responsible for his team not scoring, by either missing a FGA that is rebounded by the defense, by missing a FTA that is rebounded by the defense, or by committing a turnover. these can be estimated for a player (or counted using play-by-play)...
i then divide a player's points scored by his zero point possessions to get pts/0ptposs, which i use as a measure of offensive efficiency. so for example, this past season deandre jordan shot better overall than did kawhi leonard, a 59.1% vs. a 58.5% ScFG% (2s, 3s, and FTs). but jordan turned the ball over with 10% of his touches, leonard with just 5% of his touches. when you work out the numbers leonard had pts/0ptposs of 2.58, jordan 2.06. in other words, for each time he cost his team a team possession without scoring (and a team possession was worth on average about 1.05 pts this season), leonard scored 2.58 pts, jordan just 2.06 pts...