This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
Thanks for the reply. I tried to do do both of those things.. I traded out Kolb/Stafford for Rivers/McNabb and Felix Jones for Mike Wallace. I also reduced the amount spent on K and TD and used that to improve my WR depth. I actually like this team better now.I like every player on this team and feel good about my chances. I would like to still tweak it a bit and maybe add another low cost QB.
One last change. Now at 28/$250. I'm definitely done now.OK, just made one tweak. Still at 27/$250, but I think I like this version a little better. For real, though, I'm done now.
I'm not a betting man, but I would take that bet.One last change. Now at 28/$250. I'm definitely done now.OK, just made one tweak. Still at 27/$250, but I think I like this version a little better. For real, though, I'm done now.
The deadline to submit/edit your contest entry is 11:59pm (ET) Wednesday, September 7th.It never happens exactly at noon, always a few hours later before it locks. It'll be fun reading this thread on the 7th: obsessive hand-wringers, breathlessly reporting that it is 12:03 and they just adjusted their team!
Not noon - midnight.
Present company excepted, of course....
Wholesale changes made. Now at 26/$250.One last change. Now at 28/$250. I'm definitely done now.OK, just made one tweak. Still at 27/$250, but I think I like this version a little better. For real, though, I'm done now.
Locked and loaded with the following changes:QB $38 (2) - haven't budged on theseRB $85 (9) - these will be the triumph or downfall for me - went value, core remains the same but added lottery picks/backupsWR $80 (7) - found some value and shifted $$$ to RB's, a 3rd TE and a 4th KTE $28 (3) - liked this 3rd TE better to flex on a given week vs. other WR options at the same $$$PK $10 (4) - 4 better than 3? I'm about to find out...TD $9 (3) - dropped 2 mid priced TD's for the $3 variety to save a few $$$Total: $250 28 playersStill have 1/2 my $$$ with week 5/6 byes. Need to make it on to week 7 to have a shot. Bring on week 1!!!
I can't believe I lost the 2010 database. I could have easily updated this with last year's data.
Anyway:[QUOTE='Doug Drinen]OK, so I dumped a bunch of data and ran some regressions.
I'm going to disagree somewhat. I get what you're saying and agree that you need to look at the interaction between the positions. But I don't think that individual position analysis is meaningless. One possible strategy in this contest is to start by deciding how much many of each position you want and how much you want to spend there. The QB, Kicker and Defense positions are unique in that you can only start one. If I start by saying that I believe I need to get 10 points from my Kicker position every week, then any study (such as the one CalBear just posted) that says that 3 cheap kickers will pretty much guarantee me my desired 10 points is meaningful. Now I know i can have up to 27 roster spots left and $242 to spend. I've gotten my desired outcome from this position and eliminated one variable from the problem. You can now start to work on the rest of the contest, using this as an assumption, then at the end, you can revisit and see if removing that 3rd kicker will help/harm your team.'Marauder said:
I've done this and simulated the season, play by play, game by game, 100x but am kind of stuck with what to actually do with the data.
You can get Iggy's vote if you can find a way to add another kicker.
I thought I was done, too. Then the Garrard news broke..........
You really were done if you had him on your roster to begin with.
You've got too many Week 11 byes. I would suggest Denver D (week 6), and change out Roethlisberger to Sam Bradford. Then you've got $7 more for more kickers!
He may have had another Jag player that he felt better having on his roster with Garrard under center.
Him, no. His receivers, yes.
Just switched out Shorts. Just can't trust him with McCown, although I suppose hoping for Gabbert to play better with the younger WRs may be a good plan as well.
Sorry, wrong forum. Try the ACF.
I actually meant for this contest.Sorry, wrong forum. Try the ACF.
I think he got that... but you missed hisI actually meant for this contest.Sorry, wrong forum. Try the ACF.
I think he got that... but you missed hisI actually meant for this contest.Sorry, wrong forum. Try the ACF.
Pos Players $ SpentQB 2 $34RB 8 $78WR 8 $98TE 3 $28PK 3 $6TD 2 $6TOTALS 26 $250I agree on RB. There is no obvious Arian Foster this season. Making it very tough. I think an injury or overperformance will likely be the key to RB stability. At RB, I personally went with one stud and a bunch of longshots hoping to hit on one or two.
I just exported the draft dominator weekly scoring, which is the easy part. The harder part is trying to figure out how to weight players who might not score for your team. For instance how do you value Matthew Stafford if you already have Aaron Rodgers on your team? You expect Rodgers to handle a majority of your scoring, however to protect against injury, bye weeks, and poor performance Stafford has some value. How do you weight that against having better depth at WR? I thought going into this the optimization algorithm would be the hardest, but I found it's actually pretty easy to get something to complete with a high-value team in a few hours, it's the algorithm that scores the teams which is the hardest to get right.Edit: I've considered trying to factor in SoS into this as well but I don't have time this year. In theory you want to grab players with easier schedules as the season goes on. While SoS is already factored into projections, if you have a RB projected at 15 points for a week but is running against the #1 rush defense and you have another that's projected at 13 and is running against #32 it's probably better to go for the second guy as he's got a better chance of breaking out that week. At least I think that's the case, I need to run through some historical projections and actuals to figure if that's even correct. I haven't even figured out the sweet-spot for back-up player scoring yet so I think both of those will be a good project next summer.
You'd probably need to run simulations using some kind of pmf for player scoring.
I'd guess (but I'm not sure) that this is basically how DD comes up with its weekly player points - e.g. if you have a QB projected for 320 fantasy points for the season, you distribute that 320 points across each of the weeks, weighted by the position-specific SOS. I'm not sure how valuable preseason SOS is anyway.But this is basically where the pmfs would come in. You could derive some distributions based on past performance of players with a certain projection, and if you wanted to you could include opponent type (e.g. WRs projected to score 10-12 points playing against a team with a top-10 passing defense). Then run simulations. I think this is similar to what Drinen does for his in-season sims.All of this seems like it's far more work than it's worth, but I understand the desire to do stuff like this anyway, just for fun. If you had last year's prices and weekly player scores, could you plug them in and optimize a team for last year's contest? An open question (for me, at least) is what that team would look like (since I assume the team that won wasn't actually the optimal team). I don't have the prices and points, but I'm just curious if you could do that anyway. The prices and weekly points would be relatively easy to grab if you wanted to try.
Yeah, thats the problem i was having. My solution was to simulate each game many times so I can then get a distribution of scores for each week and through several possible season paths, then I could do as you said and have a "starter" team and work from there, but I never made it that far. I can get the score distributions just didn't have the time to complete what you did.
