Redefining TOV%

Home for all your discussion of basketball statistical analysis.
Post Reply
EvanZ
Posts: 912
Joined: Thu Apr 14, 2011 10:41 pm
Location: The City
Contact:

Redefining TOV%

Post by EvanZ »

I changed the definition of TOV% for my site. My definition is:

Code: Select all

TOV% = 100 * TOV/PLAYS
where

Code: Select all

PLAYS = TSA + AST + TOV
Anybody have issues with this? I think an assist should count as a play, or at least, an opportunity that we know the player had during a possession that he could have turned over the ball. I don't know the answer to this but my guess would be that on average the number of dribbles a player has before he shoots the ball or has an assist is roughly equivalent, so we should not neglect these plays. Seems obvious to me now.
v-zero
Posts: 520
Joined: Sat Oct 27, 2012 12:30 pm

Re: Redefining TOV%

Post by v-zero »

Makes sense from where I am, although I personally rather like TOV/(TOV + AST) if we are considering using this to represent passing in some way.
deepak
Posts: 100
Joined: Sun Apr 17, 2011 3:33 pm

Re: Redefining TOV%

Post by deepak »

What were you using before?

Your new definition more or less matches Hollinger's "Turnover Ratio", defined on ESPN.com as:

Turnover Ratio = (Turnover x 100) divided by [(FGA + (FTA x 0.44) + Assists + Turnovers]

TOV% at Basketball-reference.com removes Assists from the equation.
DSMok1
Posts: 1119
Joined: Thu Apr 14, 2011 11:18 pm
Location: Maine
Contact:

Re: Redefining TOV%

Post by DSMok1 »

Yep, that formulation is typically called turnover ratio rather than TO%, from what I have seen.
Developer of Box Plus/Minus
APBRmetrics Forum Administrator
Twitter.com/DSMok1
EvanZ
Posts: 912
Joined: Thu Apr 14, 2011 10:41 pm
Location: The City
Contact:

Re: Redefining TOV%

Post by EvanZ »

I use TOR for turnover rate (tov per 100 team possessions). So many acronyms so little time.
EvanZ
Posts: 912
Joined: Thu Apr 14, 2011 10:41 pm
Location: The City
Contact:

Re: Redefining TOV%

Post by EvanZ »

v-zero wrote:Makes sense from where I am, although I personally rather like TOV/(TOV + AST) if we are considering using this to represent passing in some way.
According to play-by-play, Chris Paul has 12 "bad pass turnovers" and 65 turnovers due to getting the ball stolen (how many of those are actually passes is anyone's guess), 4 due to traveling, 4 due to losing the ball out of bounds, 9 due to offensive fouls, and a few others due to other acts not related to passing. So it's not clear to me that TOV by itself really represents passing all that well.

For reference, Stephen Curry has 18 "bad pass turnovers". Maybe I should put that stat on the site, so people can see the breakdown for each player and make their own decisions/stats based off of it.
Mike G
Posts: 6175
Joined: Fri Apr 15, 2011 12:02 am
Location: Asheville, NC

Re: Redefining TOV%

Post by Mike G »

I prefer calling a TO a TO, rather than a TOV.
Bobbofitos
Posts: 306
Joined: Sat Apr 16, 2011 7:40 am
Location: Cambridge, MA
Contact:

Re: Redefining TOV%

Post by Bobbofitos »

Mike G wrote:I prefer calling a TO a TO, rather than a TOV.
I like the 2 letter acronym better, too.
EvanZ
Posts: 912
Joined: Thu Apr 14, 2011 10:41 pm
Location: The City
Contact:

Re: Redefining TOV%

Post by EvanZ »

Then what's a timeout called?

At any rate (pun intended), I've added a "Turnover Details" table to my site under the "Ball Handling" tab. It looks like this:

Image
Mike G
Posts: 6175
Joined: Fri Apr 15, 2011 12:02 am
Location: Asheville, NC

Re: Redefining TOV%

Post by Mike G »

Maybe "steal" should read "stolen"?

It's funny to see these Warriors and their TO broken down by type. I was just noticing that Andris Biedrins leads the league in TO% at 35.3 . This is also the 2nd highest TO% of the current Century (after Oberto, Was, 2010).

Biedrins' Usg% is a nearly-nonexistant 3.9 -- barely half of the 2nd-lowest in the league this year, and well below any of the other 100-highest TO% since 2000 (Ruffin, '05 : 6.0)

Just 1 or 2 of his 11 TO seem to be from attempting to do anything with the ball; the rest are moving picks and standing around fouls.
And he's hardly a "turnover machine". His 1.1 TO/36 is lowest on the team. That's about 1 every 4 games.
(He only goes 3 games on avg between FGA!)

What kind of TO rate or ratio makes sense to describe such a player?
v-zero
Posts: 520
Joined: Sat Oct 27, 2012 12:30 pm

Re: Redefining TOV%

Post by v-zero »

EvanZ wrote: According to play-by-play, Chris Paul has 12 "bad pass turnovers" and 65 turnovers due to getting the ball stolen (how many of those are actually passes is anyone's guess), 4 due to traveling, 4 due to losing the ball out of bounds, 9 due to offensive fouls, and a few others due to other acts not related to passing. So it's not clear to me that TOV by itself really represents passing all that well.

For reference, Stephen Curry has 18 "bad pass turnovers". Maybe I should put that stat on the site, so people can see the breakdown for each player and make their own decisions/stats based off of it.
Those would be nice numbers to have quickly available. I don't personally do anything special with TOs except split them into 'standard', 'fast break' and 'last second', and that has provided a good basis for me. If a player gets more TOs because he handles/uses the ball more then that added handling/use should show up in his offensive numbers, so I don't feel the need to fiddle more.

I too, prefer the TO convention.
Post Reply