Auction Valuation Method? (1 Viewer)

[sweetness34]

I’m looking for input on whether or not this is a stupid way to go about estimating auction values – as a starting point. I have the results (prices)of the last 3 years worth of auctions. I then sort all the prices from highest to lowest within each position (all player names are irrelevant at this point) and average them between the 3 years. Then I take my own rankings/projections and copy them straight over to lineup with the average price. E.G., The avg. price paid for the most expensive QB has been $68. So, if Peyton is my #1 ranked QB, his value will start at $68. The prices will decrease from there according to QB2, QB3, etc…. Apply this theory to all positions. I will use that as a “base” value for each player since this is factoring in the spending tendencies of my own league. Then, I can have a separate column to set my “true” value for each player by either adjusting the given value up or down from that base value.

I realize this method may be simplified to the point of being worthless and since I’ve made no mention of VBD or ADP or DD or baselines or XYZDPDQ, I should be embarrassed to consider myself a FBG.

Thoughts?

 

[TeamDingo]

It is what I do.

 

[CalBear]

What you're measuring with that method is not value, but expected cost. Expected cost is a useful bit of information to have, but you also need to look at VBD or some other value measure to decide whether it's appropriate to pay the expected cost.

My league systematically over-pays for QBs; that is, the amount of money you pay for 1 point of VBD is more for QBs than for RBs or WRs. Knowing that QBs are going to be overpriced, I know where I can spend my money more effectively, and wait for a bargain at QB.

 

[sweetness34]

What you're measuring with that method is not value, but expected cost. Expected cost is a useful bit of information to have, but you also need to look at VBD or some other value measure to decide whether it's appropriate to pay the expected cost.My league systematically over-pays for QBs; that is, the amount of money you pay for 1 point of VBD is more for QBs than for RBs or WRs. Knowing that QBs are going to be overpriced, I know where I can spend my money more effectively, and wait for a bargain at QB.

Thanks CalBear. Yeah, i knew that method wouldn't give me a true value for the player - in a VDB-type sense. I just wondered if that was a decent place to start - knowing that's how my league has drafted in the past.

 

[BigRed]

I’m looking for input on whether or not this is a stupid way to go about estimating auction values – as a starting point. I have the results (prices)of the last 3 years worth of auctions. I then sort all the prices from highest to lowest within each position (all player names are irrelevant at this point) and average them between the 3 years. Then I take my own rankings/projections and copy them straight over to lineup with the average price. E.G., The avg. price paid for the most expensive QB has been $68. So, if Peyton is my #1 ranked QB, his value will start at $68. The prices will decrease from there according to QB2, QB3, etc…. Apply this theory to all positions. I will use that as a “base” value for each player since this is factoring in the spending tendencies of my own league. Then, I can have a separate column to set my “true” value for each player by either adjusting the given value up or down from that base value.

I realize this method may be simplified to the point of being worthless and since I’ve made no mention of VBD or ADP or DD or baselines or XYZDPDQ, I should be embarrassed to consider myself a FBG.

Thoughts?

I think this was way too busy for me to even read the entire post, and offhand say it's more work than it's worth. Just my .02.

 

[CalBear]

What you're measuring with that method is not value, but expected cost. Expected cost is a useful bit of information to have, but you also need to look at VBD or some other value measure to decide whether it's appropriate to pay the expected cost.My league systematically over-pays for QBs; that is, the amount of money you pay for 1 point of VBD is more for QBs than for RBs or WRs. Knowing that QBs are going to be overpriced, I know where I can spend my money more effectively, and wait for a bargain at QB.

Thanks CalBear. Yeah, i knew that method wouldn't give me a true value for the player - in a VDB-type sense. I just wondered if that was a decent place to start - knowing that's how my league has drafted in the past.

It's a very good place to start. I would follow up by building a mock team; know that you'll want to spend a certain percentage on RBs, WRs, QBs, etc., so check your expected cost list and some sort of value-based list to develop a realistic budget for each position.

 

[beto]

What you're measuring with that method is not value, but expected cost. Expected cost is a useful bit of information to have, but you also need to look at VBD or some other value measure to decide whether it's appropriate to pay the expected cost.My league systematically over-pays for QBs; that is, the amount of money you pay for 1 point of VBD is more for QBs than for RBs or WRs. Knowing that QBs are going to be overpriced, I know where I can spend my money more effectively, and wait for a bargain at QB.

Thanks CalBear. Yeah, i knew that method wouldn't give me a true value for the player - in a VDB-type sense. I just wondered if that was a decent place to start - knowing that's how my league has drafted in the past.

I think it's a real good place to start. Once you have this baseline it will give you a frame of refernce to the ecomonics of the auction. League spending trends will vary somewhat over the years, and you will know instantly where they are varying.

 

[SelenaCat]

I do something very similar - I list out the prices straight from the previous years' auctions, but also include a +/- points per game based on my own baselines (more or less the worst backup for each position). It's easy to tell which players are values based on the amount you'd expect to pay. I tend not to pay too much attention to the actual "value" of a player, however, since it's a very fluid thing once the auction gets started.

 

[sweetness34]

I do something very similar - I list out the prices straight from the previous years' auctions, but also include a +/- points per game based on my own baselines (more or less the worst backup for each position). It's easy to tell which players are values based on the amount you'd expect to pay. I tend not to pay too much attention to the actual "value" of a player, however, since it's a very fluid thing once the auction gets started.

Thanks for your thoughts SC. I've done an auction for a few years, but i've never felt like I've nailed down the best way to prep. I'm coming to the conclusion that trying to pick/calculate/VBD-ise the "actual" or "true" value of a player is very difficult and really, not that valuable since everything is relative to your particular league/auction and each individual person's spending habits. Obviously, one needs to have a general or close approximation of where they value every player (that's where previous years data comes in), but arguing with someone whether LT is worth $78 or $72 seems like a moot (moo )point.(Then again, maybe this perspective is the root of my problem...?)

 

[Steel Dillo]

After reading this thread, I went back and created a similar spreadsheet using 4 years of past auction data.

I think it will be very helpful for our league, especially in the first third of the auction because our league uses a blind auction. All starters & backups' names are randomly drawn out of a bowl, each owner writes down a bid, then we go around the room and announce our bids. This style of auction always has three phases. Pretty fair market price for the first third of the auction, some huge overpayments when people are trying to make sure they don't get stuck without a stud in the second third of the auction and some great deals in the last part of the auction when a stud pops ups after most everyone has blown their salary cap.

Now that I have this tool, if a must have stud like LT comes out early, I'll have a good idea what its going to take to get him.

 

[Q-Bert]

I think this is the start of a very good system. Here are some tweaks I would make.

Take that list of the average price from the last three years and instead of applying it to your rankings - apply it to an ADP ranking. This will tell you roughly how much each player will go for. Now apply it to your rankings as above. In a spreadsheet set up the following columns:

Player Name........Expected Cost.........True Cost...........Value

Where Expected cost is your comparison of past results to ADP, True Cost is your comparison of past results to your ranking, and Value is True Cost - Expected Cost. Look to sign guys who have a positive number in the Value column.

This would be a very simple way to prepare for your auction. If you wanted to get more complex you could use some of the methods described in various subcriber articles for determining what the values in the True Cost column should be.

 

[Q-Bert]

Some articles on how to determine auction values.

Correcting Auction Values in the VBD Spreadsheet

The Perfect Auction

Auction Pricing

The more I think about it the more I like combining the above methods (for determining true value) with your idea of looking at your leagues past history and ADP info (for determining expected cost) to find the players that are undervalued.

 

[sweetness34]

I think this is the start of a very good system. Here are some tweaks I would make.Take that list of the average price from the last three years and instead of applying it to your rankings - apply it to an ADP ranking. This will tell you roughly how much each player will go for. Now apply it to your rankings as above. In a spreadsheet set up the following columnslayer Name........Expected Cost.........True Cost...........ValueWhere Expected cost is your comparison of past results to ADP, True Cost is your comparison of past results to your ranking, and Value is True Cost - Expected Cost. Look to sign guys who have a positive number in the Value column.This would be a very simple way to prepare for your auction. If you wanted to get more complex you could use some of the methods described in various subcriber articles for determining what the values in the True Cost column should be.

This is what I've been looking for. That was essentially what I wanted to do, but you're right (and CalBear said it too), I already had the expected cost. the true cost is what i needed to implement. thanks for explaining it so succinctly, Q.(and thanks for the links. i have those saved somewhere, but now i won't have to go searching. thx)

 

[Illegal Pants Downfield]

I take a similar approach - of course, it requires some tweaking, but I feel like it gives me a solid place to start, so I know I'm in the ballpark for any given player...

 

[stickboy]

I use a similar approach, but not to determine "value". I plot out the pricing along with the points projected for each position to determine where I want to bid. An auction allows me to pick where I want to spend, so I use this information to determine whether I should target RB #2 or RB # 5, where can I most effectively spend my auction $$. From there, I now use projections and other factors to determine who is most likely going to be RB #2. The "value" comes in finding players that will most likely outperform their consensus rank or ADP, not in auction $$. If I am targeting RB #2 and my projections say that RB #3 is a much better value, I may adjust my budget to take RB #3 and move up on some other curve.

 

[sweetness34]

I use a similar approach, but not to determine "value". I plot out the pricing along with the points projected for each position to determine where I want to bid. An auction allows me to pick where I want to spend, so I use this information to determine whether I should target RB #2 or RB # 5, where can I most effectively spend my auction $$. From there, I now use projections and other factors to determine who is most likely going to be RB #2. The "value" comes in finding players that will most likely outperform their consensus rank or ADP, not in auction $$. If I am targeting RB #2 and my projections say that RB #3 is a much better value, I may adjust my budget to take RB #3 and move up on some other curve.

This is great advice, thanks.