RealGM

i think this is the final version of the formula. number next to a player's name represents his marks. same as last time, i'm asking for constructive criticism and observations
my site won't be finished before the end of this month, due to both objective and subjective reasons, but it will be up and running before the end of may so you can play around with it during the summer and get familiar enough before fantasy-draft season starts. i'm not going to reveal tools i plan on implementing because nothing is set into stone and i might decide to leave something out at the last moment. all i can tell you is that it will have highly customizable ratings of players, teams and trades and will offer some customizing options never before seen(at least by me) on existing sites. all this and much more



formula was further improved, results are in 3rd post

I am intrigued by your theories and wish to subscribe to your news letter.

DNYLkiller, i can't say if you're being sarcastic or not:)

new results of the reworked formula, hope you like it, i loved it:):

1. paul,chris 232,25
2. james,lebron 228,75
3. stoudemire,amare 225,08
4. bryant,kobe 215,82
5. camby,marcus 211,34
6. davis,baron 206,41
7. nowitzki,dirk 200,58
8. butler,caron 196,69
9. ginobili,manu 196,29
10. marion,shawn 196,10
11. nash,steve 195,56
12. garnett,kevin 193,86
13. iverson,allen 193,36
14. ming,yao 193,33
15. billups,chauncey 192,68
16. smith,josh 191,50
17. howard,dwight 187,56
18. williams,deron 186,67
19. richardson,jason 185,81
20. kaman,chris 185,65
21. duncan,tim 185,19
22. pierce,paul 183,67
23. granger,danny 183,04
24. bosh,chris 181,57
25. lewis,rashard 181,31
26. gasol,pau 180,42
27. jackson,stephen 179,73
28. bynum,andrew 178,22
29. artest,ron 174,92
30. wallace,gerald 174,41
31. dunleavy,mike 174,32
32. wallace,rasheed 174,20
33. gay,rudy 173,55
34. carter,vince 173,10
35. allen,ray 172,98
36. martin,kevin 172,45
37. jefferson,al 172,43
38. turkoglu,hedo 172,30
39. jamison,antawn 171,63
40. iguodala,andre 171,20
41. anthony,carmelo 171,15
42. wade,dwyane 170,49
43. boozer,carlos 170,06
44. johnson,joe 169,07
45. redd,michael 167,66
46. calderon,jose 166,73
47. west,david 166,23
48. williams,mo 164,98
49. crawford,jamal 162,24
50. kirilenko,andrei 162,16
51. kidd,jason 161,40
52. stojakovic,peja 161,01
53. ellis,monta 159,91
54. miller,brad 159,50
55. miller,mike 157,65
56. odom,lamar 157,40
57. maggette,corey 157,20
58. roy,brandon 155,78
59. gordon,ben 154,27
60. hill,grant 152,93
61. mcgrady,tracy 152,27
62. dalembert,samuel 151,39
63. jefferson,richar 151,20
64. barbosa,leandro 151,11
65. howard,josh 150,65
66. o'neal,jermaine 150,04
67. ilgauskas,zydrun 148,72
68. terry,jason 148,48
69. aldridge,lamarcu 147,10
70. hamilton,richard 146,69
71. battier,shane 146,62
72. harris,devin 146,45
73. durant,kevin 146,18
74. bogut,andrew 144,22
75. bell,raja 142,25
76. arenas,gilbert 142,11
77. okur,mehmet 141,65
78. parker,anthony 140,67
79. miller,andre 140,62
80. bibby,mike 139,42
81. murphy,troy 139,29
82. biedrins,andris 138,94
83. alston,rafer 138,54
84. childress,josh 138,21
85. harrington,al 137,82
86. moon,jamario 137,03
87. chandler,tyson 136,26
88. okafor,emeka 136,23
89. gibson,daniel 136,15
90. haywood,brendan 136,09
91. martin,kenyon 134,94
92. hinrich,kirk 134,75
93. brewer,ronnie 133,37
94. felton,raymond 133,03
95. deng,luol 130,09
96. parker,tony 128,99
97. tinsley,jamaal 128,97
98. garcia,francisco 128,49
99. fisher,derek 127,94
100. mccants,rashad 127,61
101. lee,david 126,85
102. salmons,john 126,28
103. carter,anthony 126,24
104. nelson,jameer 125,31
105. williams,marvin 125,00
106. prince,tayshaun 124,85
107. udrih,beno 124,66
108. mobley,cuttino 124,33
109. randolph,zach 124,11
110. horford,al 124,11
111. smith,j.r. 123,20
112. thomas,tim 123,08
113. davis,ricky 122,70
114. nocioni,andres 122,24
115. williams,jason 122,06
116. gomes,ryan 120,54
117. ford,t.j. 120,29
118. korver,kyle 120,20
119. stevenson,deshaw 120,15
120. haslem,udonis 118,96
121. o'neal,shaquille 118,89
122. robinson,nate 117,95
123. rondo,rajon 117,87
124. marbury,stephon 116,58
125. collison,nick 116,54
126. dampier,erick 116,50
127. hughes,larry 115,62
128. wright,dorell 115,11
129. perkins,kendrick 115,03
130. kleiza,linas 114,98
131. jones,james 114,00
132. watson,earl 113,93
133. outlaw,travis 113,70
134. wilcox,chris 113,45
135. blake,steve 113,42
136. gooden,drew 112,59
137. bogans,keith 111,83
138. daniels,antonio 111,54
139. webster,martell 111,12
140. barry,brent 110,88
141. jaric,marko 110,72
142. thomas,kurt 110,30
143. szczerbiak,wally 109,90
144. turiaf,ronny 109,75
145. navarro,juan_car 109,65
146. blatche,andray 108,95
147. mcdyess,antonio 108,79
148. millsap,paul 108,70
149. farmar,jordan 108,40
150. posey,james 107,81
151. foye,randy 107,76
152. delfino,carlos 107,60
153. diaw,boris 107,08
154. cassell,sam 106,64
155. williams,louis 106,07
156. radmanovic,vladi 105,63
157. stackhouse,jerry 105,55
158. lasme,stephane 105,28
159. mason,roger 105,17
160. pietrus,mickael 105,12
161. wallace,ben 104,94
162. moore,mikki 104,58
163. barnes,matt 103,86
164. smith,joe 103,74
165. jack,jarrett 103,59
166. telfair,sebastia 103,22
167. carroll,matt 102,77
168. lowry,kyle 102,55
169. williams,sean 102,08
170. vujacic,sasha 102,06
171. bargnani,andrea 101,56
172. house,eddie 100,76
173. azubuike,kelenna 100,50
174. przybilla,joel 100,26
175. najera,eduardo 99,87
176. landry,carl 99,84
177. milicic,darko 99,83
178. jianlian,yi 99,77
179. nachbar,bostjan 99,76
180. green,willie 99,73
181. finley,michael 99,69
182. ratliff,theo 99,56
183. villanueva,charl 99,45
184. foster,jeff 99,36
185. thornton,al 98,79
186. maxiell,jason 97,48
187. evans,maurice 97,44
188. rush,kareem 97,13
189. warrick,hakim 96,48
190. knight,brevin 96,43
191. jones,fred 96,40
192. mason,desmond 95,66
193. young,thaddeus 95,51
194. nesterovic,rasho 95,33
195. noah,joakim 95,16
196. peterson,morris 94,71
197. bass,brandon 94,19
198. mourning,alonzo 94,10
199. johnson,anthony 93,70
200. scola,luis 93,67
201. bell,charlie 93,17
202. wells,bonzi 92,94
203. smith,craig 92,49
204. west,delonte 92,40
205. jones,damon 92,32
206. boone,josh 92,15
207. wilkins,damien 91,94
208. head,luther 91,88
209. varejao,anderson 91,85
210. quinn,chris 91,72
211. walton,luke 91,29
212. banks,marcus 91,10
213. diener,travis 90,85
214. ridnour,luke 89,78
215. green,jeff 89,40
216. arroyo,carlos 89,27
217. bowen,bruce 88,99
218. dooling,keyon 88,98
219. sefolosha,thabo 88,82
220. jackson,bobby 88,55
221. curry,eddy 88,45
222. thomas,tyrus 88,37
223. mohammed,nazr 88,25
224. conley,mike 88,18
225. hayes,chuck 86,76
226. williams,shawne 86,19
227. snyder,kirk 86,16
228. murray,ronald 86,13
229. daniels,marquis 85,84
230. richardson,quent 85,67
231. duhon,chris 85,37
232. pargo,jannero 85,20
233. johnson,amir 84,27
234. ariza,trevor 84,19
235. kapono,jason 83,53
236. wright,antoine 83,26
237. oberto,fabricio 83,21
238. udoka,ime 82,57
239. brown,devin 82,36
240. hayes,jarvis 82,16
241. swift,stromile 81,27
242. diop,desagana 81,07
243. blount,mark 81,06
244. powe,leon 81,02
245. harpring,matt 80,61
246. lue,tyronn 80,56
247. allen,tony 80,43
248. cook,daequan 80,38
249. songaila,darius 80,14
250. hardaway,anferne 79,77
251. frye,channing 79,10
252. williams,marcus 78,54
253. mutombo,dikembe 77,65
254. dudley,jared 77,55
255. walker,antoine 77,49
256. simmons,bobby 76,75
257. skinner,brian 76,30
258. petro,johan 76,06
259. stuckey,rodney 75,18
260. anthony,joel 75,03
261. bonner,matt 74,96
262. dickau,dan 74,95
263. mahinmi,ian 74,92
264. butler,rasual 74,79
265. atkins,chucky 74,15
266. allen,malik 73,24
267. samb,cheick 73,08
268. stoudamire,damon 72,90
269. carney,rodney 72,85
270. hilario,nene 72,79
271. smith,jason 72,38
272. marshall,donyell 72,24
273. miles,c.j. 71,48
274. young,nick 70,78
275. brewer,corey 70,30
276. ivey,royal 70,14
277. powell,josh 69,81
278. diogu,ike 69,79
279. pavlovic,sasha 69,63
280. francis,steve 69,40
281. anderson,derek 68,99
282. humphries,kris 68,95
283. mcinnis,jeff 68,64
284. ross,quinton 68,63
285. wright,brandan 68,40
286. jones,eddie 68,07
287. harrison,david 67,95
288. gill,eddie 67,92
289. newble,ira 67,16
290. webber,chris 67,11
291. cook,brian 66,87
292. giricek,gordan 66,71
293. croshere,austin 66,10
294. patterson,ruben 65,46
295. brown,kwame 64,95
296. graham,stephen 64,92
297. jones,bobby 64,51
298. buckner,greg 64,43
299. parker,smush 63,75
300. krstic,nenad 63,73
301. gelabale,mickael 63,56
302. dixon,juan 63,38
303. jackson,luke 63,35
304. williams,aaron 63,26
305. james,mike 62,94
306. powell,kasib 62,64
307. horry,robert 62,30
308. hawes,spencer 62,13
309. davis,glen 61,80
310. stoudamire,salim 61,74
311. brooks,aaron 61,66
312. owens,andre 61,66
313. pachulia,zaza 61,53
314. evans,reggie 61,46
315. barea,jose_juan 61,22
316. wright,julian 61,15
317. johnson,dermarr 60,87
318. douby,quincy 60,79
319. novak,steve 60,60
320. watson,c.j. 60,56
321. wilks,mike 60,39
322. baston,maceo 60,34
323. vaughn,jacque 60,06
324. chandler,wilson 59,57
325. fazekas,nick 59,42
326. george,devean 59,32
327. barron,earl 59,04
328. green,gerald 58,99
329. harris,mike 58,64
330. garbajosa,jorge 58,46
331. swift,robert 58,43
332. price,ronnie 58,09
333. williams,shelden 57,82
334. balkman,renaldo 57,78
335. james,jerome 57,69
336. johnson,alexande 57,48
337. doleac,michael 57,41
338. law,acie 57,29
339. boykins,earl 57,02
340. ruffin,michael 56,96
341. afflalo,arron 56,79
342. armstrong,hilton 56,71
343. cardinal,brian 56,70
344. sessions,ramon 56,24
345. hollins,ryan 55,97
346. mihm,chris 55,71
347. bowen,ryan 55,69
348. mbenga,dj 55,64
349. elson,francisco 55,30
350. voskuhl,jake 55,12

A+... you get a star for effort.

thanks 4Tay

Hard to help you with your formula, give advice or criticism, if we cant see your formula....or am I missing it somewhere?

no formula. i need you to point out players who you think are over/under-rated. i just have to reiterate that this is based on per game season averages for standard 9 cats. i don't expect there will be any major objections, but still i wanted to get some opinions

It all sounds great man. In all honesty I have not followed fantasy basketball that much to give you good feedback on who is under or overrated on your list.

Give me a pm when your site goes up though I am looking forward to take a look at it.

My take:
Overvalued:
Jamal Crawford
Dwight Howard
Daniel Gibson
Tim Thomas
Andres Nocioni

Undervalued:
Tyson Chandler
Reggie Evans
Emeka Okafor
Brandon Roy
Jeff Foster
Maurice Evans
Josh Boone
Joakim Noah
Biedrins
Al Jefferson
Rondo
Ben Wallace
Andrew Bogut
Anderson Varejao

My comments:
Big ups for greatly evaluating some high-risers of this season such as J-rich(flys under the radar all the time in fantasy land),S Jax ,Dunleavy,Hedo and Gay.The top of your list seems pretty accurate as well.IMO the biggest questionmark is the ranking of centers.Shaq and Dwight are ranked surprisingly high whereas Biedrins,Bogut,Chandler,Wallace,Varejao and Okafor are ranked surprisingly low .Looking at this I'd say that points are slightly overrated whereas turnovers are slightly underrated.

What I mostly notice about your formula is rebounds are a bit underrated,also in terms of percentages, the amount of shots taken by the player seems to have very little effect on the rankings.Also for H2H,specialists such as Reggie Evans,Jeff Foster,who dont hurt you a lot in other cats and help you a lot in one or two cats,are probably more valuable than their current ranking.

All in all great job altogether,I just wanted to give my 2 cents.

KalElen, my questions. I don't expect you to respond to these ony-by-one if you don't want to!

KalElen wrote:17. howard,dwight 187,56 - IMO far too high. He's a four -cat player and a liability everywhere else. Overrated by this formula

27. jackson,stephen 179,73 - just don't see him as a top-30 pick. He's more suited to being in the 40 or 50s.

43. boozer,carlos 170,06 - I think he deserves a bit more love. He's a great three-cat player, but also contributes partially to others. He's at least on a par with Dwight (in fantasy) IMO.

49. crawford,jamal 162,24 - too high. Same issue as with S-Jax.

59. gordon,ben 154,27 - way too high. I haven't check the stats, but I thought this was an off-year for BG? Are you over-valuing FT%?

61. mcgrady,tracy 152,27 - too high for T-Mac. Frankly, he doesn't have 'it' for fantasy.

66. o'neal,jermaine 150,04 - Small sample size, but even on per-game stats, he shouldn't have been 66th.

76. arenas,gilbert 142,11 - same as JO.

89. gibson,daniel 136,15 - looks like 3pt are being overrated in a big way. Boobie doesn't do much else.

121. o'neal,shaquille 118,89 - yeah, too high.

126. dampier,erick 116,50 - I could never take Erica this high, unless it was the WNBA draft

127. hughes,larry 115,62 - I could never Larry this high

128. wright,dorell 115,11 - my boy deserves some more love!

136. gooden,drew 112,59 - think he warranted going higher, at least his Chicago stats

209. varejao,anderson 91,85 - no biggie, but Varejao is a little better than 209 IMO.

221. curry,eddy 88,45 - no biggie either, but E-City shouldn't sniff 300.

344. sessions,ramon 56,24 - Small sample size, but he's looked like a stud!

And my observations:

nostradamus2005 wrote:My take:
Undervalued:

Tyson Chandler - he's not under-valued at all. He didn't block many shots. Dalembert is easily better!

Reggie Evans - meh, he's really a waiver-wire pickup. He will kill your FT% and he is a one-trick pony.

Emeka Okafor - once again, I use the Dalembert comparison. Meka was healthy this season, but 57% FT won't cut it...nor will decreased blocks.

Brandon Roy - I think we shouldn't overrate Roy. Everyone puts him high because he is the (true) savior of the Portland franchise, but his fantasy game won't translate as well.

Rondo - 123 is fair for Rondo. He doesn't shoot 3s and his FT% is awful. Everything else doesn't really mask those glaring holes.

Ben Wallace - I think he's done. KalElen was spot on.

of course i'll respond, to both of you. this is exactly what i wanted when i made the thread. that list is based on almost a month old stats though so, i'll generate one based on end season stats and then respond

17th ranked Dwight Howard
599 .590 0.0 20.7 14.2 1.3 0.9 2.1 3.2
88th ranked Emeka Okafor
535 .570 0.0 13.8 10.7 0.9 0.8 1.7 2.0

I dont see 71 spot difference.Yeah Howard scores 7 more points and 3,5 more rebounds but those are the easiest stats to get.He has slight advantage in blocks by a 0.4 margin but he also turns the ball over 1.2 times more per game.Also Dwight's abysmal ft% is more hurting since he has 6 more FTA per game.I'd probably say Dwight is somewhere around mid -30's whereas Emeka is around mid-60's.

How can Reggie Evans kill your ft% when he takes 3.2 FTA per game.He is a two trick pony(rebounds and steals) and the players ranked 50 spots above him dont even help you in anything at all(Bonner,Croshere etc.)I'd say since he at least helps you in two cats,he should be within 250 range.

Brandon Roy-Compare his stats to that of Hedo Turkoglu.Fg%,blocks and points are a wash.Assists and rebounds combined are a wash.Hedo has better fg% and hits 1 more 3 per game(significant difference) however Hedo has also 1.8 turnovers more per game.So they basicall cancel each other out and I'd say there arent 20 spots difference between two players.

Rondo-He takes 2.1 FTA per game,so his damage in that category is pretty minimal.His line is 10.5 ppg 5.1 apg 4.2 rpg a very good 1.7 spg and boasts very good fg% at 49.5 for a pg with a 1.9 turnover rate which is very reasonable for a pg.I'd say he is more like around early 100's than 120's

Ben Wallace is washed up but he is still a valuable fantasy contributor.His %'s are horrible but he doesnt have a lot of attempts so he aint hurting his owners that much.He still gets 7.4 rpg 1.7 bpg 0.9 spg 0.9 turnovers per game and he plays as a center(valuable position) so I'd say he is more around 120's than 160's

sorry guys but i just don't have the time to respond properly before sunday. hope you won't lose interest by then

@ nostradamus2005 & Young_Star11 thanks for taking the time to respond guys, i appreciate it. formula is created to ensure that none of the cumulative stats(ppg, apg, rpg...) is under/overrated because i have extremely good way of ranking them(dare i say perfect way), my focus when making the thread was more on %s because they are extremely hard to accurately mark. here are the stats of an average nba player for the season 07/08:
3ptm 0,54
rpg 3,58
apg 1,82
spg 0,61
tpg 1,16
bpg 0,41
ppg 8,27
fg% 0,451 fgm 3,08 fga 6,84
ft% 0,746 ftm 1,56 fta 2,09

my rankings use these numbers to calculate value of a unit of production for every category(for 7 cumulative stats). if you reduce number of players considered, by taking out bad players or players who had small pt(in a word scrubs) those stats would go up, but this list was created to include all players who played this season

there is a popular belief that goes roughly like this: 1bpg=1spg=-1tpg=1 3pg=2apg=2rpg=3ppg (i saw countless people asking advice about their team on nba's forum playing in point leagues with those weights). but as you can see that's far from true.

turnovers are not underrated, you can see that many scrubs average more then 1 tpg so that reduces negative effect of 3+tpg players. i was surprised by that fact too, but it's the fact nonetheless. 62 top 100 player average 2+ tpg while just 5 average 1-

rebounds are not underrated either. as you can see from 3.58 league average too many players contribute in rebounding in order for it to carry great weight. while rebound is more then twice as valuable as a point at the same time it's almost 7 times less valuable then a three pointer and almost 9 times less valuable then a block

3ptm are not overrated, they are 2nd most valuable category after blocks. they are a little top heavy so you can get an impression that they are easy to come by but if you go a little deeper(beyond 100-150 ranking) it becomes clear that there are more players contributing to steals then to treys

points are definitely not overrated, they are the least valuable category(per unit of production) and as such carry the least points (again per unit of production or per 1 ppg if you prefer)

finally attempts and makes are calculated in %s calculations. d-how for instance is almost the worst ft% shooting player though there are many players shooting lower ft% but his high fta numbers keep him down. btw worst ft% shooter by my formula is shaq

nostradamus:

crawford is where he is thanks to the fact the he is big contributor to ft% and 3ptm, well above average in assists(assists are not as valuable on average as 3ptm but they are even more top heavy so solid assist contributor is no small thing); he is also solid in ppg and spg without hurting too much on tpg. i'm not saying he should be drafted here, i'm saying his production this year ranked him here

d-how ended the season ranked 30th. i don't know why he fell 13 spots in last month, i don't own him so i didn't follow him closely but i think you'll both agree that sounds about right for this season(he made significant improvements from last season). he's ft% might not be penalized enough though and i'll definitely look into it

gibson finished 102nd; he was above average in several cats and huge in 3ptm, but his ft% brings him 2nd highest mark. i guess there still might be problem in that department since his ranking really seems a little too high

tim thomas finished 119th, i don't think that's too high for his stats; his contributions are spread out but maybe he too got better marks for ft% then he should have

nocioni,andres finished 120th. i don't know what to tell you, i don't see problem with him here

chandler finished 87th, he is a 3 cat player; with his blocks slipping this year, there's no way he can go higher

evans finished 357th, he can't get too much credit for 7.54 rpg since it's 2nd cheapest cat and his steals aren't impressive either, but he is one of the worst players in ft% effect, he really hurts you there, he is almost 30% below the league average and he take more then 3 attempts which makes him 88th in the league - there are not many players who can pull such negative influence up. if you have couple of top ft% players or decide to tank ft% and play very deep league he can be useful, but if you are just solid in ft% he can single-handedly kill you there

okafor finished 86th, and if you compare him to chandler it's clear that he is where he belongs, he is just slightly better. as for huge gap between him and d-how, i'd say that the problem in d-how's rank is made clear, but not with okafor's

roy finished 58th, i don

KalElen wrote:young_star
boozer finished 46th, i think he is vastly overrated by many people, he is 3 cat contributor, but 2 of those cats are the cheapest 2(ppg and rpg) and he doesn't even dominate them; his fg% contribution is 4th in the league by my formula, but he doesn't help much in 4 cats, steals and turnovers about cancel each other out and you are left with just those 3; that's barely top 50 in my book(even without formula)

gordon finished 60th, he is similar to s-jax, better in few cats, worse in few others; funny thing is that i'll probably take him above s-jax, that's how much i don't trust jackson, but numbers are numbers

hughes finished 136th, he got surprisingly many points on ft%, i'll have to look into that too

dorell wright finished 125th, i don't get this one, based on what should he go higher? his main qualities 1- block and 1- turnover; nice %s can't help to much with so few attempts

eddy curry finished 243rd, based on fg% effect, 54.56% on 9.22 attempts is no small thing

obviously there are still some issues with %s that i'll have too look into(i have a pretty good idea about what i think is the cause of the problem), but overall i think it's a good list and therefore a good formula. i just want to thank you both one more time for taking the time to go through the list

Boozer would be top-10 (or close to) in both rebounds and points, no? Is that not dominating? His FG% on so many attempts is also dominant. 3 assists for a center and 1.2 steals compared to 2.6 turnovers is still a net positive IMO. I'll admit blocks and free-throw turnovers are shabby and of course he doesn't hit threes, but I would've thought that he'd be a perfect foil to LeBron in the late second or early third.

Gordon is still a little debatable. I've just checked his numbers, and he decreased his TO to about 2.1, a bit more palpable. I still think you are overvaluing his 3PTM and FT%, no matter how valuable they are.

Yeah, Dampier is overrated.

Hughes probably gets overrated because he can steal, hit a few threes and have a good impact on FT% Debatable at 136, perhaps he should be at 150, but it's alright.

I'm a big fan of Dorell. His steals and blocks are nice and so are the turnovers. His percentages are great for a guard-eligible player, and he rebounds quite well. A case to go a little higher, despite him only playing 44 games.

We all know about Eddy Curry. Points (barely) and FG% are his only cats, seriously. He doesn't help anywhere else, in fact he hurts you. 243 and falling...

I think it is a good formula, you went with league averages, which I have been trying to incorporate into my ranking system...but I never have enough time.

You might have picked up on (in my above explanations) my penchant for stats relative to his position.

You went with the stats of an average NBA player, I was thinking maybe (if you had the time of course) if you could kind of create a system for each position, but with similar weightings (if you get my drift).

But apart from some questionnable (IMO) rankings, I think you have done a great job this season KalElen. RealGM has definitely appreciated your input in season 2007/08.

KalElen - when using averages for analyzing fantasy basketball worth, it seems to me, it would make better sense to use averages based on "fantasy-worthy" players. Based on observations from some roto leagues I've been in, for the cumulative stats (in standard size 12-team leagues), you are looking at base averages of around 1 3pg, 15 ppg, 5.7 rpg, 3.4 apg, 1 spg, 0.7 bpg, 1.9 topg, which would affect the weights of the categories. (You did state that if you take away players with small pt or bad stats, the numbers go up, but not only would numbers go up, but the weighings change as well. In the end rankings/strategies/worth change due to league size. In addition, consideration has to be made for positions used in a league, which lends credence to Youngstar's suggestion of using averages based on position. (Unless, your formula takes into consideration of the number of players needed by position and position eligibility. I've played in starting one C leagues, and that makes a noteworthy difference in valuing players compared to starting two C leagues.) Though, I think if you present a ranking, you have to make assumptions, such as standard size league with standard starting positions.

Also, in regards to Nostradomus' post, I think any ranking formula won't take into account h2h strategy worth. Hence 2-cat players such as Reggie Evans should never rank high under a formula that considers all categories. h2h strategists would need to rank players specifically based on strategy.

hamncheese i agree with your observations about weights changing with exclusion of scrubs, i had that in mind all along(my comments you referred to, were intended to clarify this, but i guess they were too obscure). there is one simple problem though and that is to recognize fantasy worthy players in the 1st place. easy solution would be to try to rule out as many scrubs as possible based on their low minutes per game and games played numbers. my plan is to give those options to the user: to choose minimum of minutes per game and/or games played for a player to get onto the list, that way users in deeper leagues can have lower requirements and users in shallower leagues can have higher requirements and both of them can leave out most of the scrubs. but i wanted to test out the formula as soon as possible(before most fbb players stop visiting boards due to season ending). i have to work on several other aspects of the site too, so i've been forced to postpone realizing the above mentioned option. before i test this i can't be sure that on its own this solution is enough, but it will definitely be a step in the right direction. 3s will lose some of their value in that case because they are top heavy and so will assists, but values of blocks, points and rebounds shouldn't change as much

as far as strategies go, the fact remains that same players have different values to different teams using different strategies(tanking certain cats, % management), but formula can't take any of that into account because it is meant to be based on all cats. i do plan on devising some interesting tools in this department though, so stay tuned...


finally about positions issue Young_Star11, hamncheese brought up. a block is a block no matter where it comes from. if a guard records a block in addition to everything he is expected to do his rank will be higher because of it, but that 1 block will not be any more valuable to you then the block that adonal foyle gets because it won't help you win category any more then foyle's. guard getting that block is rewarded by getting uncharacteristically high number of points in that category. if a center gets you a 3 it will help his ranking but it will still be just a 3. if you look at 2 notable 3 point shooting centers troy murphy and okur you can see that they both have relatively high ft%, low fg% and low blocks so you can't reward them for producing odd stats with extra points when they don't produce common ones. i know that good pgs are more valuable then good sgs, but that's mostly due to the fact that majority of assists comes from limited number of top pgs and that is already covered by formula. also some leagues don't differentiate pgs from sgs, but rather play 2 guards, 2 forwards etc. yet another problem would be multy-positional eligibility. take diaw from couple of years ago, he was sg-sf-pf-c eligible, so his ft% was good for a c, but bad for a sg, his rebs were ok for a c, but huge for a sg, his lack of 3s didn't hurt if you played him as a c, but did hurt if you played him at sg. which eligibility do you take into account? if you have a solution to those problems i'll be glad to listen, but i don't think there's a way to do it without complicating things mathematically beyond recognition. i believe that a 3 is a 3 no matter where it comes from, so position eligibility has no place in an objective formula based on stats. i'm thinking about inserting "intangibles" category to every ranking which might reward multy-positional eligibility, in some way(to be determined later, feel free to offer suggestions) if user so desires

as for boozer, being top 10 in a cat is not dominating, being 1st and almost 25% in front of the runner up(camby in front of smoove in blocks) is dominating. boozer is not 1st, let alone dominant, in any cat. his ast+stl-to is a net positive, but not big positive. ignore a cat or 2 and he looks much better, but the end result is based on 9 cats and 3 cat contributors just can't go too far overall especially if he specializes in 2 cheapest cats

gordon's 3s are not overrated by the formula. if we exclude bottom 100 players they get less valuable, but in these settings they are that valuable to get s-jax and gordon to where they are

dorell wright averages 4.95 rpg, 1.36 apg, 0.66 spg, 0.66 tpg, 0.91 bpg, 7.89 ppg, 48.8 fg% on 6.66 fga, 82,8 ft% on 1.57 fta; clearly he doesn't hurt you anywhere, but he almost doesn

and i almost forgot, thanks Young_Star11, i just wish my writing for rotofreak went better, but soon after the start of the season i couldn't come up with any interesting ideas for the articles. still it's been an interesting experience

Young_Star11 wrote:But apart from some questionnable (IMO) rankings, I think you have done a great job this season KalElen. RealGM has definitely appreciated your input in season 2007/08.

KalElen - Nice responses. I really admire the effort you are undertaking. The point about positions and why they matter is less about the stats they produce (as you say, a block is a block), but more about distribution and being able to rank effectively based on position. It's general conventional wisdom that centers/power forwards are the more difficult positions to fill, followed by point guards. Based on one ranking I have, out of the top 120 players, roughly 30 players are PG or PG/SG eligible, 45 are SG, SF, or SG/SF eligibile, 17 are SF/PF or PF eligible, and 28 are C or PF/C eligible, based on Yahoo eligibility. If you are in a standard 12-team league needing to start 1 PF and 2 C's, then quality of big man production is "scarcer", hence the value of position should go up, because even if you can get same value from SG or SF or even stats, you may not be able to play that player. This would affect the value of a player. As for how to take that into account, I don't know entirely, but I might group players in the four ways I've done above - (1) PG, PG/SG (2) SG, SF, SG/SF (3) PF, SF/PF, (4) C, PF/C.

i see your point, but what would you do with those groups? if you want to add position scarcity into the equation you need to reward some positions, but by how much? it's easy to rank positions amongst themselves, grouped for example in a way you suggested, but how much weight do those rewards carry. using the formula i can tell you how many rebounds is a block worth, but how do you calculate by how many rebounds(or anything else) is a c more valuable then a sf?
if i really tried i could come up with some ways that include stats more then others, but at some level it will come down to arbitrary choosing a number and that would negate all objectivity that formula accomplishes in judging statistics
there are 3 more major problems:
-league settings vary so much that for some of them standard yahoo settings make no sense and as such must not be included in the "basic" formula (as i mentioned earlier, some leagues don't make distinction between pgs and sgs but play gs, different leagues start different number of players on different positions and util position was originally intended to solve this problem you're talking about by giving gms a chance to play additional players of those more common positions but add further complexity to different settings problem)
-where do you put diaw with sg/sf/pf/c or charlie bell pg/sg/sf etc. if you want to cover all possibilities you get too many groups which raises complexity of calculations
-not everyone is playing on yahoo and my goal is to create a tool universally helpful to fantasy nba players

ideas that might help without eliminating formula