PDA

View Full Version : Advanced stats



save the knicks
02-05-2010, 11:22 AM
So what is everyones take on things like eFG TS% and the four factors?
Are these metrics really more precise?
Should eFG or TS replace FG% why or why not?
Can anyone out there with a firm grip on this stuff explain it or point me in a the right direction?

Chronz
02-05-2010, 04:15 PM
Great thread, Im really interested as to why its taking the general population so long to catch up, this stuff has been around forever now.

ChiSox219
02-05-2010, 04:21 PM
Great thread, Im really interested as to why its taking the general population so long to catch up, this stuff has been around forever now.

Many baseball fans still reject sabermetrics even though just about every team uses them. :shrug:

kozelkid
02-05-2010, 06:42 PM
Great thread, Im really interested as to why its taking the general population so long to catch up, this stuff has been around forever now.

Very true.

As for the thread, here goes. Fg%, while certainly not a bad stat, does not tell the whole story, what I mean by whole story is this. Consider a 3pt shooter who shoots 33% from 3 (and for sake of my point ONLY 3's) and a big man who shoots purely 2's at 50%. You'd think the big man is WAY more efficient because of the fg%, when in fact, they are identical. You have one player making a third of his 3's so he is scoring 1 point per possession and the other player making half his 2's which is also a point per possession. So they are identical. Fg% doesn't take into account the point value of different shots. Ts% does. Not only that, but ts% also takes into account fts which is just as important. A player might get fouled and that will neither help or hurt his fg%, but if he makes both fouls, that should certainly indicate he is more efficient.
So the formula for ts% is points/2(shots + (.44*fta)). And league average for guards is around .53. I believe big man is around .55. Now you might ask why .44 in the equation. Simple reason, the creator of the stat calculated that an average of 12% of the fts are technical fouls, flagrant fouls, clear path fouls, fouls beyond the arc, and ones etc. Reasoning behind this is that these fts don't take away from a team's possessions total.

As for efg%, this is basically ts% minus the fts. I generally don't use it since I consider fts just as important... efg%= (fg+.5*3pt)/fga

I applaud you for making this thread and it will help you understand the game that much more for using such stats. These are more of the basic advanced stats (haha, oxymoron). Anywho, if you have anymore questions on advanced stats, feel free to ask them or pm me. Also chisox219 is a big fan of APM and probably knows more about them than me. Personally I'm not a fan, but they are considered another sort of advanced statistics.

Verbal Christ
02-05-2010, 07:01 PM
really old article, but rockets GM darryl morey explains it best IMO (not that any of you wouldnt know this stuff, but you get to pick his brain a bit here) kinda long so i attached just a few morsels to see if you want to continue.

http://www.houstonpress.com/2007-11-01/news/rocket-science



Contrary to popular belief, statistics do not lie. At worst, the numbers are merely unwilling accomplices forced into an arena to which they don't belong. The true masters of deception are the people who, consciously or otherwise, flaunt their misleading stats to continue perpetuating half-truths and outright lies.



Buoyed by the discovery of a kindred spirit and his dominance in the emerging world of fantasy sports (that's right, fantasy junkies, you too can become a real-life GM — just as soon as you graduate from Northwestern and MIT, of course), Morey set out on a path he hoped would allow him to make his mark on the professional sports landscape.



The mechanisms of change had been set in motion. Sabermetricians and statistical analysis geeks were popping up all over the country, and their message was slowly beginning to weave its way into the minds of the sporting public.



Says Morey: "I think the bottom line, though, is that — where basketball gets really interesting and exciting — is that you can't isolate things down into one or two simple concepts. Because while it's true you can isolate the team level to that, you can't do it at the player level. For example: Does a guy hit a shot because he's a great shooter? Because he got a great pick? Because the pass was good? You can't isolate that perfectly with any sort of objective evidence. So you're always developing a picture of a player and you're always integrating the two. Whereas in baseball there are fewer variables and you may be able to isolate it better."



"From a traditional box score I'd say there's only 20 to 30 percent that's useful relative to some of the more advanced stuff you can measure. That's why we have our — we call it our army — people who track stuff leaguewide"



"You can see basketball as a series of mini-games," he says. "When ten players are on the floor, they play for anywhere between 30 seconds and three minutes and until you sub someone in, you can see basketball as a mini-game between that five-on-five set. And you're trying to win each mini-game throughout the game, and you can then basically make conclusions on whether or not a player's winning the mini-games they're in throughout the game."

Shady66
02-05-2010, 08:22 PM
Very true.

As for the thread, here goes. Fg%, while certainly not a bad stat, does not tell the whole story, what I mean by whole story is this. Consider a 3pt shooter who shoots 33% from 3 (and for sake of my point ONLY 3's) and a big man who shoots purely 2's at 50%. You'd think the big man is WAY more efficient because of the fg%, when in fact, they are identical. You have one player making a third of his 3's so he is scoring 1 point per possession and the other player making half his 2's which is also a point per possession. So they are identical. Fg% doesn't take into account the point value of different shots. Ts% does. Not only that, but ts% also takes into account fts which is just as important. A player might get fouled and that will neither help or hurt his fg%, but if he makes both fouls, that should certainly indicate he is more efficient.
So the formula for ts% is points/2(shots + (.44*fta)). And league average for guards is around .53. I believe big man is around .55. Now you might ask why .44 in the equation. Simple reason, the creator of the stat calculated that an average of 12% of the fts are technical fouls, flagrant fouls, clear path fouls, fouls beyond the arc, and ones etc. Reasoning behind this is that these fts don't take away from a team's possessions total.

As for efg%, this is basically ts% minus the fts. I generally don't use it since I consider fts just as important... efg%= (fg+.5*3pt)/fga

I applaud you for making this thread and it will help you understand the game that much more for using such stats. These are more of the basic advanced stats (haha, oxymoron). Anywho, if you have anymore questions on advanced stats, feel free to ask them or pm me. Also chisox219 is a big fan of APM and probably knows more about them than me. Personally I'm not a fan, but they are considered another sort of advanced statistics.

One of the best posts iv ever seen on PSD.

save the knicks
04-13-2010, 07:37 AM
awesome, glad i checked out my old posts.

Lots of good info hear

Thanks