slinslin
Welcome to Amareca
Oh ok I didn't read that. Didn't they have a more static system before? That you always had the #1 and #16 picks in the rounds when you had #1 in the first and #16 in the second? At least that's how I remembered it.
Fantasy Draft Thread
- Thread starter Chris_Sanders
- Start date
Oh ok I didn't read that. Didn't they have a more static system before? That you always had the #1 and #16 picks in the rounds when you had #1 in the first and #16 in the second? At least that's how I remembered it.
Chaplin
Better off silent
No, this site has always been like it is this year--only two rounds "snake" at a time, after that the draft order is re-randomized. In some ways, it's good, in others, not.
elindholm
edited for content
Maybe two years ago or more, but not last year. This is exactly the way it was last year. The "randomness" is part of the logic. I think it's more fair than always having the same draft order in rounds 1, 3, 5....
slinslin
Welcome to Amareca
Ok maybe it is only this year that I realize that I would prefer to be in a different draft position.
At least I got a good combination in the 2nd league #3, #30, #34.
At least I got a good combination in the 2nd league #3, #30, #34.
Chaplin
Better off silent
But then again, doing a "snake" draft, as they do in Yahoo and just about every other fantasy site, gives you a clear idea of what your picks' values are.
Personally, I like this way myself. I wish we could make it a keeper league and then run it next year like the real NBA--I'm in an NFL keeper league, and it's probably more complicated than it needs to be, but it's pretty fun.
OP
OP
Had to go with Lebron there. In the preseason it is his passing that has sold me on him.
thegrahamcrackr
Registered User
I was wondering where he would land
Errntknght
Registered User
I'm not an interested party but I remember a way we used to pick 'sides' for softball and volleyball games. A draft if you will for two teams. A coin was flipped to see who chose first and the winner chose one player, then the other captain picked two players then the winner picked two and so on. The way this would work for a multiteam draft would be to have a single random assignment of draft order. As always the first round is picked in that order then the next TWO rounds picked in the reverse order. Following that there are two rounds in the original order then two in the reverse order, et cetera.
Randomizing after every two rounds is fair because everyone has an equal chance of gaining an advantage but if you want to be fair and reduce the effects of chance as much as possible then the method I've outlined is better. I suspect one could even make a case for doing one round in the original order then three rounds in the reverse order etc.
(O migosh, thats two in two days!)
Randomizing after every two rounds is fair because everyone has an equal chance of gaining an advantage but if you want to be fair and reduce the effects of chance as much as possible then the method I've outlined is better. I suspect one could even make a case for doing one round in the original order then three rounds in the reverse order etc.
(O migosh, thats two in two days!)
OP
OP
Ugh...
That was a hard pick. I might have taken Rasheed Wallace if not for the fact he is definitely on the downside and Portland has been playing him at SF. His rebound numbers are due to go down.
Meanwhile, Nene is pretty much guaranteed minutes, gets nice steal numbers, big rebounder, and can play the middle.
That was a hard pick. I might have taken Rasheed Wallace if not for the fact he is definitely on the downside and Portland has been playing him at SF. His rebound numbers are due to go down.
Meanwhile, Nene is pretty much guaranteed minutes, gets nice steal numbers, big rebounder, and can play the middle.
In the meantime we lost Gorilla again.
Joe Mama
Joe Mama
wierwolf
Koolaid Guzzler
I hope you are having a better turnout then us in the 2nd League. We started @ 1030 and still on Pick 6. 
OP
OP
We are into round 3 now.
Kirilienko was a very good pick. I almost took him, but I needed to shore up my center position.
Kirilienko was a very good pick. I almost took him, but I needed to shore up my center position.
What is the best advice you would give a first time player in this league? Would you pick based on need or the best player available?
For example, if you have already picked 2 PF/C's with your first two picks and the best player available is Ben Wallace...do you take Wallace or try to fill another position (PG/SG) with the pick?
For example, if you have already picked 2 PF/C's with your first two picks and the best player available is Ben Wallace...do you take Wallace or try to fill another position (PG/SG) with the pick?
elindholm
edited for content
Randomizing after every two rounds is fair because everyone has an equal chance of gaining an advantage but if you want to be fair and reduce the effects of chance as much as possible then the method I've outlined is better.
What? To reduce unfair effects of chance, you want to avoid any system that arbitrarily imposes a systematic pattern. If there is an advantage to choosing 1st and 32nd versus 16th and 17th (or if the advantage is the other way), you want to randomize which teams benefit from that advantage in subsequent round-pairs.
What? To reduce unfair effects of chance, you want to avoid any system that arbitrarily imposes a systematic pattern. If there is an advantage to choosing 1st and 32nd versus 16th and 17th (or if the advantage is the other way), you want to randomize which teams benefit from that advantage in subsequent round-pairs.
slinslin
Welcome to Amareca
There usually isn't a player available who is cleary better than a couple of other good choices.
Personally I think that it is tougher to get good guards in the later rounds than good forwards but that's just based on my personal experience.
If I already had 2 FCs and had to chose between a G,F or C with roughly the same fantasy production I would rather pick a guard.
On the other hand it wouldn't be a total waste to pick another FC or C because 3 FCs can cover all your minutes at center and like half your minutes at forward.
Personally I think that it is tougher to get good guards in the later rounds than good forwards but that's just based on my personal experience.
If I already had 2 FCs and had to chose between a G,F or C with roughly the same fantasy production I would rather pick a guard.
On the other hand it wouldn't be a total waste to pick another FC or C because 3 FCs can cover all your minutes at center and like half your minutes at forward.
Errntknght
Registered User
Eric wrote, "To reduce unfair effects of chance, you want to avoid any system that arbitrarily imposes a systematic pattern."
I suppose that's true as stated but it isn't true that a well chosen, as opposed to arbitrary, systematic pattern increases the 'unfair' effects of chance. I had in mind when I made up those patterns that the difference between a player and the next best player decreases the further you go into a draft. This is because one would expect that the amount of basketball talent in a population would follow something like a normal (bell shaped) distribution.
I put that to a test using simulated data, the upper half of the normal distibution - meaning I generated 512 normally distributed random data points, then used the top half for 16 rounds of drafting by 16 teams. The results are below.
The first line gives sort of a baseline - it shows the advantage the first team picking would have over the last if you set up a draft order then simply repeated that order for 16 rounds. The second line shows the effect of re-randomizing the order for each round - a big improvement over the first case. But notice that well chosen patterns of Reversing the draft order have an stronger effect than randomizing alone and, indeed, mixing in randomization with the pattern makes it less effective in some cases - it's less pronounced an effect than I expected. There is no absolute best pattern as it depends on the underlying distribution of basketball talent. I tried 'top third of Normal' and got about the same result however.
Average of 5000 trials each case,
FRpattern=F Rndm=N Hi-Low range=20.24
FRpattern=F Rndm=Y Hi-Low range= 7.91
FRpattern=FR Rndm=N Hi-Low range= 5.63
FRpattern=FR Rndm=Y Hi-Low range= 5.45
FRpattern=FRRF Rndm=N Hi-Low range= 5.21
FRpattern=FRRF Rndm=Y Hi-Low range= 5.25
FRpattern=FRRRFF Rndm=N Hi-Low range= 4.53 <- best
FRpattern=FRRRFF Rndm=Y Hi-Low range= 5.08
FRpattern=FRRRRFFF Rndm=N Hi-Low range= 4.73
FRpattern=FRRRRFFF Rndm=Y Hi-Low range= 4.83
FR pattern means a pattern of using a set draft order in Forward and Reverse order. So FRRF means use it one round in Forward order then two rounds in Reverse order then once again in Forward order. The whole pattern repeats until the draft is done. Rndm=Y means that you re-randomize the order after the pattern is complete and before it is repeated.
I suppose that's true as stated but it isn't true that a well chosen, as opposed to arbitrary, systematic pattern increases the 'unfair' effects of chance. I had in mind when I made up those patterns that the difference between a player and the next best player decreases the further you go into a draft. This is because one would expect that the amount of basketball talent in a population would follow something like a normal (bell shaped) distribution.
I put that to a test using simulated data, the upper half of the normal distibution - meaning I generated 512 normally distributed random data points, then used the top half for 16 rounds of drafting by 16 teams. The results are below.
The first line gives sort of a baseline - it shows the advantage the first team picking would have over the last if you set up a draft order then simply repeated that order for 16 rounds. The second line shows the effect of re-randomizing the order for each round - a big improvement over the first case. But notice that well chosen patterns of Reversing the draft order have an stronger effect than randomizing alone and, indeed, mixing in randomization with the pattern makes it less effective in some cases - it's less pronounced an effect than I expected. There is no absolute best pattern as it depends on the underlying distribution of basketball talent. I tried 'top third of Normal' and got about the same result however.
Average of 5000 trials each case,
FRpattern=F Rndm=N Hi-Low range=20.24
FRpattern=F Rndm=Y Hi-Low range= 7.91
FRpattern=FR Rndm=N Hi-Low range= 5.63
FRpattern=FR Rndm=Y Hi-Low range= 5.45
FRpattern=FRRF Rndm=N Hi-Low range= 5.21
FRpattern=FRRF Rndm=Y Hi-Low range= 5.25
FRpattern=FRRRFF Rndm=N Hi-Low range= 4.53 <- best
FRpattern=FRRRFF Rndm=Y Hi-Low range= 5.08
FRpattern=FRRRRFFF Rndm=N Hi-Low range= 4.73
FRpattern=FRRRRFFF Rndm=Y Hi-Low range= 4.83
FR pattern means a pattern of using a set draft order in Forward and Reverse order. So FRRF means use it one round in Forward order then two rounds in Reverse order then once again in Forward order. The whole pattern repeats until the draft is done. Rndm=Y means that you re-randomize the order after the pattern is complete and before it is repeated.
elindholm
edited for content
FRpattern=FR Rndm=N Hi-Low range= 5.63
FRpattern=FR Rndm=Y Hi-Low range= 5.45
Based on these two entries, it appears that the system hoops.ws uses is slightly preferable to the simple "snake" system.
Thanks for the analysis. There are times when I wish my knowledge of programming languages didn't stop with Fortran.
FRpattern=FR Rndm=Y Hi-Low range= 5.45
Based on these two entries, it appears that the system hoops.ws uses is slightly preferable to the simple "snake" system.
Thanks for the analysis. There are times when I wish my knowledge of programming languages didn't stop with Fortran.
OP
OP
What the heck is going on with Def Jam?
elindholm
edited for content
I have no idea. It's the same guy who was Birdsville last season -- who went deep into the playoffs, as you may recall -- so I figured he was a good bet to have his act together this year. Obviously I was wrong.
Chaplin
Better off silent
Well, if he doesn't show up and we end up skipping him, there's a good bet that he's going to be a "dead" team all season long.
Chaplin
Better off silent
Originally posted by elindholm
Thanks, Chaplin, rub it in.
Well, Eric, what can I say, except "I told ya so"?
fordronken
Registered User
Eric, you seem to be making, or desiring a couple "please have breakout year" picks.
Similar threads
Latest posts
-
Ossenfort's crap drafting continued. The Athletic All Rookie Team- Latest: Crimson Warrior
-
-
-
-
-
-
Staff online
-
Dback JonDoing it My Way -
Brian in MesaAdvocatus Diaboli