Im pretty sure Hollinger sets the value of an assist at 1/3 of a made bucket because the passer is responsible for a third of the sequence it takes to make an assist, the act of getting open and making the shot fall on the assistee. That is debatable because what about in drive and kick situations where the playmaker is the one who created the play by drawing an extra defender but thats not what Im goin to get into. My point is seeing as how hollinger sets value of an assist like this, wouldnt it make sense to do the same with fg/m that resulted from an assist. Set the value of assisted fg/m to 2/3 of a made fg For example Carmelo Anthony is assisted on 60% of his made buckets whereas Kobe Bryant is assisted only on 40%. By this logic it would increase the value of Kobes fg/m. I have no idea if PER accounts for this but if it doesnt my question then is why not?
I wouldn't really call it a detriment to a player if they're able to play effectively off the ball. Carmelo is also able to create shots for himself, so it's not a situation where he's overly dependent on the assist to score pointts, it's just that he's able to score without holding the ball. George Karl has preached that since he came to Denver, and Carmelo used to be a huge culprit of the stagnant ball.
<div class="quote_poster">Chronz Wrote</div><div class="quote_post">Im pretty sure Hollinger sets the value of an assist at 1/3 of a made bucket because the passer is responsible for a third of the sequence it takes to make an assist, the act of getting open and making the shot fall on the assistee. That is debatable because what about in drive and kick situations where the playmaker is the one who created the play by drawing an extra defender but thats not what Im goin to get into. My point is seeing as how hollinger sets value of an assist like this, wouldnt it make sense to do the same with fg/m that resulted from an assist. Set the value of assisted fg/m to 2/3 of a made fg For example Carmelo Anthony is assisted on 60% of his made buckets whereas Kobe Bryant is assisted only on 40%. By this logic it would increase the value of Kobes fg/m. I have no idea if PER accounts for this but if it doesnt my question then is why not?</div> Nope, it doesn't account for this, since he only uses official NBA statistics to calculate PER. The formula would instead use assist% for the team instead of the individual player. But like og15 says, there is value in being able to move without the ball and get the team easy catch and score opportunities. Often, that's a product of the offensive system put in place, but some players just know how to read defenses, make the right cuts, and get easy buckets better than others. Movement off the ball often can indirectly create scoring opportunities for teammates as well. That's one of those "intangibles" that PER doesn't account for. Anyways, I have tried calculating PER for last season using this extra information. I posted it somewhere on this board a while back. I'll repost it here: ---- Here's an article on calculating PER: http://basketball-reference.com/about/per.html Here are the changes I made for the modified PER: <ul> [*]Replaced tmAST/tmFG (an estimate for the percentage of a player's field goals that were assisted) with the player's actual assisted% from the 82games player page. [*]Replaced tmPace with the team's pace while the player was on the court [*]Replaced the VOP formula (league wide points scored per possession) with a more precise version using actual number of possessions the players played, derived from 82games On/Off page. Instead of 1.034, it turns out to be around 1.074. [/list] Here's the top 30 players by the modified PER, with a limit of 500 minutes played in the season. The ast% modification seems to be the biggest factor. Note that the formula for estimating team pace used in the PER formula (fga+0.44*fta+to-orb), which I show below as estPace, overestimates it by about 1% in general (that's why VOP was off). With Arenas on the floor, for instance, the Wizards actually play at a much faster pace, even though below estPace and pace are the same. <div class='codetop'>CODE</div><div class='codemain'><font size="1"><br/><font color="Navy"> FullNameTmMP estAst% estPace PERast%pace PER2 PERdiff</font><br/>1LeBron JamesCLE 3361 53.6% 89.828.1 32% 89.3 29.9 1.8<br/>2Dwyane Wade MIA 2897 55.7% 91.627.5 34% 91.3 29.3 1.8<br/>3Kobe Bryant LAL 3273 58.0% 90.928.1 44% 91.1 29.0 0.9<br/>4Allen Iverson PHI 3101 55.1% 92.725.9 24% 92.8 28.4 2.5<br/>5Dirk Nowitzki DAL 3086 50.0% 87.828.2 51% 87.1 27.9 -0.2<br/>6Kevin Garnett MIN 2960 59.4% 88.826.8 61% 88.1 26.8 0.0<br/>7Elton Brand LAC 3096 57.1% 91.726.6 60% 91.5 26.2 -0.3<br/>8Steve NashPHO 2801 63.5% 95.823.2 23% 95.5 25.5 2.3<br/>9Chauncey BillupsDET 2928 66.1% 86.823.4 36% 85.7 25.2 1.8<br/>10 Yao MingHOU 1946 58.3% 8825.7 62% 88.7 25.0 -0.7<br/>11 Gilbert ArenasWAS 3383 51.2% 92.323.8 37% 92.3 24.6 0.7<br/>12 Paul Pierce BOS 3087 58.1% 92.223.6 48% 92.2 24.1 0.5<br/>13 Shaquille O'NealMIA 1805 55.7% 91.624.3 58% 91.4 24.0 -0.3<br/>14 Tim DuncanSAS 2787 57.4% 88.523.0 50% 87.4 23.7 0.7<br/>15 Chris PaulNOK 2809 53.7% 8922.1 21% 89.6 23.6 1.6<br/>16 Chris BoshTOR 2751 52.9% 91.223.1 49% 90.1 23.5 0.4<br/>17 Pau Gasol MEM 3132 57.8% 86.422.6 50% 85.4 23.3 0.6<br/>18 Tracy McGrady HOU 1746 58.3% 8822.0 39% 87.9 23.2 1.2<br/>19 Vince CarterNJN 2908 67.8% 89.821.6 50% 89.7 22.7 1.1<br/>20 Shawn MarionPHO 3268 63.5% 95.823.5 76% 95.5 22.7 -0.8<br/>21 Manu Ginobili SAS 1816 57.4% 88.522.3 49% 88.9 22.5 0.2<br/>22 Ray Allen SEA 3022 55.1% 91.622.3 50% 90.6 22.5 0.2<br/>23 Tony Parker SAS 2715 57.4% 88.520.8 33% 88.2 22.5 1.7<br/>24 Zydrunas IlgauskasCLE 2283 53.6% 89.821.9 59% 87.8 21.9 0.0<br/>25 Carmelo Anthony DEN 2940 62.4% 93.922.1 63% 93.5 21.7 -0.4<br/>26 Gerald WallaceCHA 1897 58.0% 93.421.3 53% 93.2 21.7 0.4<br/>27 Andrei KirilenkoUTA 2606 64.6% 8820.6 57% 86.1 21.5 0.8<br/>28 Mike JamesTOR 2921 52.9% 91.219.8 30% 90.5 21.2 1.4<br/>29 Michael ReddMIL 3127 59.4% 91.221.3 59% 91.3 20.9 -0.4<br/>30 Carlos Arroyo ORL 59651.4% 88.418.8 25% 86.9 20.7 1.9</font></div> You can see that this version of PER is a bit more guard friendly, since their baskets generally aren't assisted as much. This makes sense, since PER is an offense-oriented stat, and guards/perimeter players are usually more important to an offense than bigs (who, conversely, are typically more important on defense). LeBron James, Dwyane Wade, Kobe Bryant, and Allen Iverson all vault ahead of Dirk Nowitzki for top PER honors. With this version, Steve Nash now looks like a legitimate MVP candidate statistically.
