APBRmetrics

[tempe85]

Sorry if I might sound a little uneducated here, I'm sort of new to the basketball statistical game. I have been wondering how people calculate offensive efficiency. I did quite a few google searches and never was able to come up with a definitive answer. The reason I want to know is because I've wanted to calculate the Phoenix Suns Offensive Efficiency and compare them to teams of the past. I'm almost sure they rank within the top 3 all time (drawn from basketball-reference.com) however I haven't been able to find out because I don't know the formula.

Sorry if this is so basic to a lot of you but it would be much appretiated. Thanks.

By the way the top two offensive efficiency's in NBA history are:

1. 1987 Los Angeles Lakers 112.3
tie 1992 Chicago Bulls 112.3

[jkubatko]

[tempe85]

[S.K.]

Check out the stats page at www.knickerblogger.net/stats/ - that'll give you a better context for the ratings.
The Suns league-leading total is indeed outstanding. They're currently sitting at 112.1 (with a league average of 102., 9.3 points above average. The Mavericks last year put up a 110.1 mark with a league average of 100.8, also 9.3 above average. However, if you divide the totals instead of subtracting them (which gives Dallas proper credit for succeeding in an environment with fewer points), the '04 Mavs take a slight advantage, 109.2 to 109.0 Add in the fact that the season isn't quite over (and the possibility of an 'expansion effect') and I'd say that the Suns fall slightly short.

On the other hand, when you consider that John Hollinger called last year's Mavericks "far and away the best offense in the NBA's modern era", second place ain't too shabby.
_________________
No books - no articles - no website.
Just opinions.

Ill-informed opinions.

[tempe85]

[Kevin Pelton]

82games' per-possession numbers are not the same as the estimates. That might be because they're more accurate -- someone posted recently that the current method overestimates possessions because offensive rebounds that aren't secured by an individual player (i.e. a ball rolls out of bounds and goes back to the offense) aren't counted -- but you can't compare those numbers to the ones on Basketball-Reference or KnickerBlogger.net (which is updated through a day or two ago -- the date on there is the only thing out of date).

[kjb]

The simplest formula for estimating offensive efficiency yourself is:

100 x (pts / (fga - orb + tov + .44 x fta))

[bchaikin]

using offensive points per team possession as an indicator of efficiency as of last sunday i have phoenix as having the 5th best offense since the 1977-78 season:

off
pts
/poss team (def pts/poss) W-L

1.139 CHI 9192 (1.031) 67-15
1.138 BOS 8788 (1.079) 57-25
1.136 LAL 8687 (1.049) 65-17
1.136 CHI 9596 (1.005) 72-10
1.134 PHO 0405 (1.057) 56-17 as of 4/3/05

[HoopStudies]

Following the methodology in Basketball on Paper (Chp 3), looking at the increment above the league average and using the formula for possessions in there, Phoenix's offense this year ranks as the best of all time.

Given that there were rule enforcement modifications this year, this makes some sense. It's easiest to set records when you can push the limits of rules a bit.

I haven't tried yet to account for slacking off, but I think that the Suns' O would fall to 2nd or perhaps stay first if you accounted for that (see Chapter 12).

Regardless of whether they're first or 5th or 8th, they're historically good. And they did it using both a modification of style and getting talent. If you don't believe style and coaching matter, look at this team for something hard to explain. Yeah, they were going to be better anyway, but not this much better.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.

[bchaikin]

as of last week (4/3/05) the suns had a per game point differential of +7.2 pts/g. that historically translates to a W-L % of about .728, or a full season W-L record of 60-22 (their actual current W-L record translates to about 63-19 over 82 games). simulation shows them with an average W-L record of 59-23...

but if using simulation i replace each of the starters on the suns (stoudemire for 36 min/g, marion 40min/g, johnson 40 min/g, richardson 36 min/g, nash 32 min/g) with their 03-04 stats, simulation shows them going just 39-43. replacing each individually with their 03-04 stats (the other 4 starters using their 04-05 stats) here's how many games worse the team is per average 82 games playing the above mentioned min/g:

-5.0 a.stoudemire
-6.0 sh.marion
-5.5 j.johnson
-2.5 q.richardson
-1.0 s.nash

that adds up to a 20 game difference over just one season, meaning each of those players "improved" by those many games in one season. the question is why...

many would point to steve nash as being the reason, but here's some food for thought. joe johnson and amare stoudemire are 3rd to 4th year and 2nd to 3rd year players. this is typically when you would see the most improvement in a player, specifically a player that's played starter's minutes since day one in the league. granted the 5 to 5.5 game improvement is large, but again this is when you would expect to see the most improvement in a player...

plus almost all of stoudemire's improvement has come with his 2pt shooting. his rebounding, shot blocking, how often he commits fouls and gets steals, his touches/min (0.9 in 03-04, 1.0 in 04-05) are all pretty much the same in 04-05 as in 03-04. but the improvement in 2pt FG% is very large, almost 9%. also his turnovers per touch have dropped from 9% to 7%. joe johnson's improvement seems to have come mostly in his 3pt shooting, its at 47% now but he was a career 32% 3pt shooter coming into this season...

also in the past shawn marion had played almost as good as he has this season back in 00-01, so there is some precedence for his playing this well. same with quentin richardson - he shot as well in 01-02 with the clippers as he is shooting now with the suns...