Calling Auction Sharks... (1 Viewer)

[RamMan]

I've now run several scenarios of the VBD, PD, and DD. I also ran the latest "On Line" VBD (2007 version) for auction values. Each of these programs has produced auction values all over the place. For example, DD has Peyton's auction value as $95 in a $200 cap league (that's crazy). The VBD generated sheet has Peyton valued at $53, which is more sane, but the same VBD has SJax at $58. The "auction adjusted" On Line VBD sheet has Peyton valued at $81 and SJax at $71. OY! My league is 10 teams, 14 roster spots, start 1-2-3 (WR/TE)-1-1 and we do get 6 pts. for TD passes. $200 cap.

My question is, can anyone make a case for why Peyton would be worth essentially the same as SJax? Moreover, does anyone think Peyton is worth between 25% - 40% of your cap? Personally, I've got him valued at around $42 and SJax valued at around $80. My league does tend to bid up RB's (and roughly half of the Top 10 RB's will be kept this season, making supply of RB studs rather scarce) so I have adjusted SJax's value accordingly. But, without those factors coming into play, can you make a case where Manning is worth the same as SJax?

 

[Willowdoc]

Although DD increases the value of the top QBs in 14 team leagues (it should) no way should it calculate out to those numbers. What baselines are you using? Wrong baselines are probably the most common reason for inappropriate auction values in DD.

 

[BroadwayG]

Can't make a case here. I've always had to heavily modify auction values.

 

[BigRed]

The answer to all those questions is no.

 

[amphibianbri]

Are you using Maurile's auction method and tweaking it according to how many players actually have value over $1 in your auction? If you are using JOe's secret formula or some other VBD only, your values will get all messed up. With 140 players being auctioned off, I would assume you would have 40+ that only get a $1 bid. Make sure that the positional baselines in Maurile's Auction Based add up to about 100 if you want real accurate dollar values. Good luck!

 

[CalBear]

I've been talking with Bruce Henderson about auction pricing in DD, and have discovered a few interesting facts.

1) The top players at each position have multipliers to increase their auction value. (You can see them in the auction.csv file in the DD folder). That's fine, they probably do command a premium. The problem is, those multipliers are applied to the raw point totals, rather than to the VBD number. So you may have a situation where Manning is projected to score 300 points, with 100 VBD points, and SJax is projected to score 200 points, with 100 VBD points, but the multiplier will give Manning 250 "New VBD" points, while SJax will only get 200. You can edit the auction.csv file if you want to use different multipliers (or just set them all to 1.0 if you like). (Quit and restart DD afterwards).

2) Another factor used in computing auction price is "number of expected starts." The #1 QB statistically starts more games for your fantasy team than the #12 QB (because with the #12 QB, you'll probably play matchups, while the #1 QB will start all games). Including expected starts in the calculation increases the slope of the curve for QBs. (There may wind up being an option in DD to modify this behavior).

Check out Maurile's article on auction pricing.

You should also examine your baselines, as they have a huge impact on auction values. I think by default DD will use "Maurile's Auction Method", which is close to what I use, but might be different than what you're expecting.

Personally, I think auction pricing should be directly proportional to VBD, and that multipliers should be applied to VBD instead of to raw points.

 

[RamMan]

Although DD increases the value of the top QBs in 14 team leagues (it should) no way should it calculate out to those numbers. What baselines are you using? Wrong baselines are probably the most common reason for inappropriate auction values in DD.

Thanks for your reply. I imported the fantasy points projections from PD, which I have set to "Joes Secret Formula" as a baseline. Should I try one of the other baseline methods (ie "Worst Starter Method")?

 

[amphibianbri]

FYI, my league is a 10 teamer too and with Mauriles Auction Method S Jax is at about 71 and Peyton is in the low to mid 40's

 

[CalBear]

Although DD increases the value of the top QBs in 14 team leagues (it should) no way should it calculate out to those numbers. What baselines are you using? Wrong baselines are probably the most common reason for inappropriate auction values in DD.

Thanks for your reply. I imported the fantasy points projections from PD, which I have set to "Joes Secret Formula" as a baseline. Should I try one of the other baseline methods (ie "Worst Starter Method")?

Joe's Secret Formula is definitely going to give you screwed-up auction values. Try Maurile's Auction Method.

 

[RamMan]

Although DD increases the value of the top QBs in 14 team leagues (it should) no way should it calculate out to those numbers. What baselines are you using? Wrong baselines are probably the most common reason for inappropriate auction values in DD.

Thanks for your reply. I imported the fantasy points projections from PD, which I have set to "Joes Secret Formula" as a baseline. Should I try one of the other baseline methods (ie "Worst Starter Method")?

Joe's Secret Formula is definitely going to give you screwed-up auction values. Try Maurile's Auction Method.

Thanks...I'll give it a go and let the board know what I came up with.

 

[Warpig]

ABSOLUTELY NOT!

For example; I've been in a 12 team auction league for the past 10 years. $100 salary cap. The top RB's will easily go for $50 to $65 each! Peyton (even though he's the #1 QB) might go for $20 - $22...even that is high!

 

[Willowdoc]

Although DD increases the value of the top QBs in 14 team leagues (it should) no way should it calculate out to those numbers. What baselines are you using? Wrong baselines are probably the most common reason for inappropriate auction values in DD.

Thanks for your reply. I imported the fantasy points projections from PD, which I have set to "Joes Secret Formula" as a baseline. Should I try one of the other baseline methods (ie "Worst Starter Method")?

CalBear explained this well. Absolutely use Maurile's baseline method. I will add that I set the baseline for kickers to 1 and ignore the price of the top ranked kicker.

 

[RamMan]

Although DD increases the value of the top QBs in 14 team leagues (it should) no way should it calculate out to those numbers. What baselines are you using? Wrong baselines are probably the most common reason for inappropriate auction values in DD.

Thanks for your reply. I imported the fantasy points projections from PD, which I have set to "Joes Secret Formula" as a baseline. Should I try one of the other baseline methods (ie "Worst Starter Method")?

Joe's Secret Formula is definitely going to give you screwed-up auction values. Try Maurile's Auction Method.

Thanks a million CalBear. I changed the baseline to Maurile's Auction Method and calm has settled over the land. Peyton's valued at $57 (still too high for my taste) and Sjax is at $69 (still too low for my taste). But, much much better. I'll try some of the other suggestions as well to furhter tweak the numbers.

 

[inca911]

Like calbear mentioned, Maurile's method incorporates several multipliers contained in a .csv file that are based on historical data, and also does an adjustment for projected starts. These multipliers impact the raw math unless you go into the csv file and reset the multipliers to a value of 1. The use of projected number of starts also duplicates setting raw projections for performance. I have a linear projection file that you can use that only cares about your positional baselines and projected scoring. The spreadsheet is at the bottom of this link, and incorporates a one-line fit for simplicity although the article talks about two line fit (just too hard to automate a two line fit):

http://subscribers.footballguys.com/2006/0...e_johnson01.php

Using standard VBD scoring system and projections, I have SJax at $76 and PManning at $57.

Remember that what players go for historically in your league isn't what VBD should be telling you. VBD should tell you true value for players, and you exploit the differences to win your league.

 

[RamMan]

Like calbear mentioned, Maurile's method incorporates several multipliers contained in a .csv file that are based on historical data, and also does an adjustment for projected starts. These multipliers impact the raw math unless you go into the csv file and reset the multipliers to a value of 1. The use of projected number of starts also duplicates setting raw projections for performance. I have a linear projection file that you can use that only cares about your positional baselines and projected scoring. The spreadsheet is at the bottom of this link, and incorporates a one-line fit for simplicity although the article talks about two line fit (just too hard to automate a two line fit):

http://subscribers.footballguys.com/2006/0...e_johnson01.php

Using standard VBD scoring system and projections, I have SJax at $76 and PManning at $57.

Remember that what players go for historically in your league isn't what VBD should be telling you. VBD should tell you true value for players, and you exploit the differences to win your league.

Outstanding inca....thanks very much.