Auction Cheat sheets and tiers

[detlef]

Forgive me if I asked this before and got a good answer that I've since forgotten, but I'm curious about something.

 

OK, so I've always understood the nature of tiers is that it's a grouping of players who, while the first in the tier is more desirable than the last, they are, none the less, a group from which you'd essentially be happy with any. At least, more so than dropping to the next. That, all things being equal in a snake draft, if you're looking at the last guy in a teir at one position and a couple of guys at the top of tiers at other positions, you may want to grab the last guy. That sort of thing.

 

Foster/Rice/McCoy and Rodgers/Brees/Brady being perfect examples.

 

With that in mind, and given that the market should be an excellent representation of this, you would think that auction values for players within a certain tier would follow. That is, there should be a reasonable drop-off from the last of one tier and the beginning of the next. At the same time, you shouldn't expect to see a very significant drop-off from one to the next within a given tier.

 

In other words, I could see auction values to be a bit of a fail safe to see if the tiers are truly accurate. I mean, what says, "I'd basically be happy with any of these players" like having their projected salaries very close from top to bottom. Would you be willing to pay about the same for Adrian Peterson and Jamaal Charles? If yes, then you'd think they both belong in the same tier.

 

However, at least after tier one, nothing could be further from the case.

 

The first QB in tier two is valued just barely less than all of the QBs in tier one yet more than 2x as much as the last QB in tier 2.

 

The top WR in teir 2 is barely less than the worst guy in tier one and 4x what guys at the end of that tier (all of whom, btw, are priced at a value that means you are basically not going to get them). Meanwhile, the last 4 guys in tier 2 are valued very closely to most of the next tier.

 

Same with tier 3 for RBs. At the top, guys cost about what the last guys in tier 2 cost and more than double what the guys at the end of that tier cost (all of whom cost about what the first 7 guys in tier 4 cost).

 

It essentially contradicts the entire nature of tiers.

 

Now

1) I dig the site and appreciate the articles and insight.

2) I'm not looking for someone to draft my team for me and always use these rankings as a jumping off point.

 

None the less, it does seem odd.

Edited August 13, 2012 by detlef

[DMD]

I understand what you are saying and let me answer it like this:

 

1. All auction values are merely general guidelines since there are so many variables at play and auction values can actually change after each player is selected and that position is short one more player. It's all about the money still in play in the league versus demand.

 

2. Within a tier the players should be more or less similar and the drop to the next tier demonstrates a somewhat notable decrease in expectations or an increase in risk. Tiers are highly subjective and more of an art than a science to be sure. But even within a tier the players will be arrayed in an order of preference and towards that end the top one you would be willing to pay a little more than the bottom name. The reality is that the players will not go for those exact amounts anyway.

 

3. The auction values for players are based on reality and are extrapolated from real auctions that have occurred. They do not follow where a player is far over bid nor does it follow where players are very underbid. Consider them more like sort of an average and therefore a guideline.

[delfamdelfam]

I feel the same way. For instance in tier two Steve Smith is at 16% but Percy Harvin is at 5%, 11% is pretty big for two guys in the same tier, and IMO both these guys should be valued around the same.

 

In the auction I'm doing right now Harvin is currently at 10%, now by TheHuddle's rankings I'm almost overpaying by double, but I'd much rather have Harvin on my team than Smith(think he'll have a better year), but then the issue is if I think Harvin is better than Smith do I pay the 16%.

 

I won Harvin in the auction for 10.5%, which might seem like I overpaid but compared to guys like Smith in the same tier, it's a bargain. This is where I get confused, am I getting a bargain or did I overpay by double?

[SecondString]

I understand what you are saying and let me answer it like this:

 

1. All auction values are merely general guidelines since there are so many variables at play and auction values can actually change after each player is selected and that position is short one more player. It's all about the money still in play in the league versus demand.

 

2. Within a tier the players should be more or less similar and the drop to the next tier demonstrates a somewhat notable decrease in expectations or an increase in risk. Tiers are highly subjective and more of an art than a science to be sure. But even within a tier the players will be arrayed in an order of preference and towards that end the top one you would be willing to pay a little more than the bottom name. The reality is that the players will not go for those exact amounts anyway.

 

3. The auction values for players are based on reality and are extrapolated from real auctions that have occurred. They do not follow where a player is far over bid nor does it follow where players are very underbid. Consider them more like sort of an average and therefore a guideline.

 

Makes sense...so your percentages = AAV's, and do NOT necessarily align with he Huddle's ranking, rating, or opinion of the player(s)?? But the tiers are broken down according to the Huddle's opinions of the players? Yes?

[detlef]

I understand what you are saying and let me answer it like this:

 

1. All auction values are merely general guidelines since there are so many variables at play and auction values can actually change after each player is selected and that position is short one more player. It's all about the money still in play in the league versus demand.

 

2. Within a tier the players should be more or less similar and the drop to the next tier demonstrates a somewhat notable decrease in expectations or an increase in risk. Tiers are highly subjective and more of an art than a science to be sure. But even within a tier the players will be arrayed in an order of preference and towards that end the top one you would be willing to pay a little more than the bottom name. The reality is that the players will not go for those exact amounts anyway.

 

3. The auction values for players are based on reality and are extrapolated from real auctions that have occurred. They do not follow where a player is far over bid nor does it follow where players are very underbid. Consider them more like sort of an average and therefore a guideline.

 

Thanks for taking the time.

 

Though, I'm curious, by the way you're describing them, your auction "cheat sheets" are really the auction equivalent to an ADP list. That being, that they're based on what guys are going for in actual auctions rather than what you think they should go for.

 

However, the players are still ranked in the order of your cheat sheets, which are anything but a reflection of ADP.

 

In other words, what you're suggesting is impossible.

 

It also bears mention that Huddle auction values are way off of any mock that I've done. Top QBs are going for well over what the huddle projects. Guys like B Marshall, Dez Bryant, Miles Austin, and plenty others going for way, way more. D McFadden and M Lynch go for far less. These are just a few and they hold consistent for virtually every mock that I've done.

Edited August 13, 2012 by detlef

[NavinRJohnson]

I don't know much about Auction drafts, but I'd like to throw in 2 cents about Tiers that may help. I think Tiers are pretty meaningless unless they are relative across positions. In order for this, the Tiers would have to be determined based on your points projections for each player and Value Based Drafting concepts. For example, there should never be a Tier 1 Kicker....the top kicker may start at Tier 4 depening upon your natural point breaks in VBD. I hope this makes sense. It's hard to put in words, but it should be relatively easy to see when you have your rankings on a spreadsheet.

 

Man this sounds geeky.

[Pip's Invitation]

You can't look at any one source/valuation as a guide. Auction dynamics are a tricky thing. They are different in every league, so many leagues will end up looking nothing like what is projected on The Huddle (or anywhere else). Don't worry about whether The Huddle thinks you overpaid. If you paid close to what you thought the guy was worth, be happy.

 

Where tiering really helps in auctions is alerting you to when guys from a certain tier at a certain position are almost gone. The last guy left in the first few tiers at any position is usually the subject of a bidding war, and you don't want to get involved in that.

[detlef]

I don't know much about Auction drafts, but I'd like to throw in 2 cents about Tiers that may help. I think Tiers are pretty meaningless unless they are relative across positions. In order for this, the Tiers would have to be determined based on your points projections for each player and Value Based Drafting concepts. For example, there should never be a Tier 1 Kicker....the top kicker may start at Tier 4 depening upon your natural point breaks in VBD. I hope this makes sense. It's hard to put in words, but it should be relatively easy to see when you have your rankings on a spreadsheet.

 

Man this sounds geeky.

 

I couldn't agree less. Tiers, within positions, should have nothing to do with others. The only reason why there shouldn't be a tier for kickers is because it should be your last pick. But at least with regard to other positions, I do think it's important to group blocks of players into tier within their position.

 

That doesn't mean a tier 1 TE is as important as a tier 1 RB, mind you. That's another thing entirely.

[DMD]

They are all done "by hand" by me but I do consult with what has happened. They are not merely calculated auction averages, they are our rankings with percentages.

[NavinRJohnson]

That doesn't mean a tier 1 TE is as important as a tier 1 RB, mind you. That's another thing entirely.

 

 

Well, I guess in my mind I believe it should because that's exactly what I was talking about. I can tell if a RB is projected to score x amounts of points more than a RB 20 spots below him, I know his value relative to other RBs.

 

The imporant this is that your cheat sheet speaks to you on Draft Day in a language you understand. Especially after a few beers!

[Big Country]

The imporant this is that your cheat sheet speaks to you on Draft Day in a language you understand. Especially after a few beers!

 

 

 

I hate when that happens as I usually get into an argument and lose.

[detlef]

They are all done "by hand" by me but I do consult with what has happened. They are not merely calculated auction averages, they are our rankings with percentages.

 

Still doesn't make sense. Let's put it this way.

 

Let's say that I did not intend to do any research at all. Just pay my money to the huddle and draft exactly the team they say I should. Follow the cheat sheets like they're the gospel. After all, technically, that should be a sound strategy, even if it's not fun for those of us who want to put the time in.

 

When I look at your cheat sheets or top 200, it is going to tell me to pick some guys much earlier or later than what ADP says or even other projection websites. AJ Green, for instance. It is also going to tell me that, despite what ADP says, Jordy Nelson should be grouped with the top WRs and Brandon Marshall should be a lumped with the 3rd tier. Once again, something that is certainly not reflected in ADP.

 

So, the tiers and the overall ranks reflect where you think guys should be ranked. Why stop at auction values? Why not say, "If I were drafting your team for you, I'd be willing to pay X for Steve Smith and X for Percy Harviin".

Edited August 13, 2012 by detlef

[delfamdelfam]

Still doesn't make sense. Let's put it this way.

 

Let's say that I did not intend to do any research at all. Just pay my money to the huddle and draft exactly the team they say I should. Follow the cheat sheets like they're the gospel. After all, technically, that should be a sound strategy, even if it's not fun for those of us who want to put the time in.

 

When I look at your cheat sheets or top 200, it is going to tell me to pick some guys much earlier or later than what ADP says or even other projection websites. AJ Green, for instance. It is also going to tell me that, despite what ADP says, Jordy Nelson should be grouped with the top WRs and Brandon Marshall should be a lumped with the 3rd tier. Once again, something that is certainly not reflected in ADP.

 

So, the tiers and the overall ranks reflect where you think guys should be ranked. Why stop at auction values? Why not say, "If I were drafting your team for you, I'd be willing to pay X for Steve Smith and X for Percy Harviin".

 

Yeah I don't get it much either. If you look under projected stats, Steve Smith is projected to get 250 and Harvin is projected for 244. So then why in the auction tier sheet is Smith 16% and Harvin 5%? Doesn't make much sense. If i did just buy thehuddle and use it as gospel, how am I suppose to value a guy like Harvin in an auction?

 

Same with Jordy Nelson and Antonio Brown? Both projected at 245 points, but in auction Jordy at 18% and Brown at 4%. What explains the big gap?

Edited August 13, 2012 by delfamdelfam

[DMD]

You should never take any auction values as gospel or even more than a guideline. I am working on trying to address some shortcomings with auction values but it may take until next year. The problem with assigning values is that the reality is that they are dependent on the total dollars in a league, how many teams there are in that league, how many rosters spots per team and how many starters in each position and then a working knowledge of player value both within a position and between all positions. There are a lot of factors involved and they would be unique to each league. They are made further more challenging considering that you need to factor in what sort of people are in your auctions and how they generally pursue building teams, preferences and guys they really like.

 

Here is an old auction:

 

http://www.thehuddle.com/x11/articles/sofa-auction.php

 

Tiers being applied to an auction would need to either be created figuring all those variables above or just done simply by tiering out the top 3's, then to the 12th in each position (in a 12 team league) and again at 24th, etc. so you could track how many starters are gone.

 

The only think that matters in an auction is not the auction values on your cheatsheet but your budget that you should create before hand.

 

You need to decide what sort of team you want to build since that directly impacts what you are willing to pay for a player.

 

http://www.thehuddle.com/classics/04_auction_bidding.php

 

 

Because you have to know what you are willing to pay in order to know if you should pay whatever the player is going for. Budgeting is key and critical and has to be your guiding light in an auction. If you do nothing else - make a budget.

 

http://www.thehuddle.com/classics/04_auction_budget.php

 

That allows you to decide where to spend your money and how much you can pay, If you overspend on a guy, it has to come from some other area. If you underspend, you get to pay more for something else.

 

Bottom line - never get tied up in auction values other than as a rough assessment. Next year I hope to design something that will still be a guideline, but will be more customized. Make your budget and reassess what you are doing every player that you get.

[flemingd]

So then why in the auction tier sheet is Smith 16% and Harvin 5%? Doesn't make much sense. If i did just buy thehuddle and use it as gospel, how am I suppose to value a guy like Harvin in an auction?

I think you're taking it a bit too literal. He's cutting his tiers along the same lines as he did for snake draft, but auctions tend to be "tierless" since anyone can add a buck. When you draft Smith in the 4th, you draft him on your slot at 4.10 even if you believe he was worth the 4.02. In an auction, though, exists the forum for multiple people to reflect their extra desire for him and effectively use "anyone's" 4th round pick to try to get him. Think of it as a trade-up. They are tossing a few extra bucks (their 10th rounder) to move from 4.10 to 4.02, knowing they now have less money (no 10th rounder) to get their backup QB with and will end up with Ponder (in the 14th) for $1 bidding instead of Big Ben.

[delfamdelfam]

I don't see how it's too literal. If both are projected for the same points, their shouldn't be an 11% difference between them, that's more than a few bucks.

[SecondString]

I don't see how it's too literal. If both are projected for the same points, their shouldn't be an 11% difference between them, that's more than a few bucks.

 

Yeah, I can't get my head around it either, but not gonna worry about it. I assign my own projections to each player (what I think they will go for) based on published AAV's and what I know about my league, and do my own rankings accordingly. I wouldn't think of using this or anyone else's to do my draft, but I can understand that others may want to be able to grab that sheet and run wih it. I think I might blow my mind trying to make sense across the board of that cheatsheet, but that doesn't mean I can't get some ideas from the tiers and the way they are structured.

Edited August 14, 2012 by SecondString

[detlef]

You should never take any auction values as gospel or even more than a guideline. I am working on trying to address some shortcomings with auction values but it may take until next year. The problem with assigning values is that the reality is that they are dependent on the total dollars in a league, how many teams there are in that league, how many rosters spots per team and how many starters in each position and then a working knowledge of player value both within a position and between all positions. There are a lot of factors involved and they would be unique to each league. They are made further more challenging considering that you need to factor in what sort of people are in your auctions and how they generally pursue building teams, preferences and guys they really like.

 

Here is an old auction:

 

http://www.thehuddle...ofa-auction.php

 

Tiers being applied to an auction would need to either be created figuring all those variables above or just done simply by tiering out the top 3's, then to the 12th in each position (in a 12 team league) and again at 24th, etc. so you could track how many starters are gone.

 

The only think that matters in an auction is not the auction values on your cheatsheet but your budget that you should create before hand.

 

You need to decide what sort of team you want to build since that directly impacts what you are willing to pay for a player.

 

http://www.thehuddle...ion_bidding.php

 

 

Because you have to know what you are willing to pay in order to know if you should pay whatever the player is going for. Budgeting is key and critical and has to be your guiding light in an auction. If you do nothing else - make a budget.

 

http://www.thehuddle...tion_budget.php

 

That allows you to decide where to spend your money and how much you can pay, If you overspend on a guy, it has to come from some other area. If you underspend, you get to pay more for something else.

 

Bottom line - never get tied up in auction values other than as a rough assessment. Next year I hope to design something that will still be a guideline, but will be more customized. Make your budget and reassess what you are doing every player that you get.

 

First off, I only mentioned the "take it as gospel" thing as a point of discussion. Mostly because that is somewhat implied. If someone had never done an auction, they might hope that an auction cheet sheat would be a great companion to see them through the draft. If a guy is available for less than what the sheet says, you buy him, if not, you don't.

 

And, frankly, because the other cheat sheets could basically be used as "gospel". Case in point, I had an employee last year who was invited to join a league at the last minute and had absolutely no idea about any of it. I gave her my cheat sheet (which follows the same structure as the Huddle's) a short list of rules to live by (take your kicker and D last, etc) and just said, "cross of names and take the highest available. All things equal, take the guy in the highest tier unless you've already got a bunch at that position".

 

And she ended up with a very solid team and made the play-offs.

 

Further, I understand that there's a myriad variables that effect auction values, but two of the ones you mentioned you already account for. You make two lists, one for 10 player, another for 12. Secondly, in terms of overall budget, you list %, so that would seem to work that out. As far as the rest, unique scoring or roster sizes as well as how your particular league values certain players/positions would also undermine the effectiveness of any cheat sheet. So the grain of salt is not unique.

 

I guess my point is, I think you truly missed the mark with these, especially when compared to the other valuable tools here. As I've pointed out, if they're intended to be a declaration of what you think a good price would be for these players, it fails, because it's not at all consistent with your own tiers and implied drop-offs. How do you explain the fact that there's often little or no drop-off to auction values from one tier to the next and sometimes massive ones within a given tier?

 

Further, I really don't buy your bit about values being based on what's going on in auctions, because I've seen nothing in any mocks I've done that even closely resembles what those cheat sheets are saying. If you go by Huddle values, there are certain players or groupings of players that you will either always or never get (and that's even allowing for some wiggle room. I'm talking about being prepared to pay 2x or 1/3 as much as what others seem prepared to spend. I'd love to see what auctions you're looking at that returns those values and, more importantly, how you had a meaningful batch of data when you first published the list earlier in the summer.

 

Honestly, what I think would be a more useful tool, would be one like the article you wrote about who's getting picked too early and late. Go mine the web for AAVs and provide commentary on who you think is over-priced and who you think is going too cheaply. Something like that. Either that or ignore AAV in your auction sheets the same way you ignore ADP in your cheat sheets.

[delfamdelfam]

I agree, I've used huddle cheat sheets for redraft non-auctions and as detlef has said, you can use it as a gospel. Now I know more about redraft so I can make my own decisions.

 

With auctions, I'm fairly new to them, so I figured I could use it just like the re-draft, kinda like a gospel. What I've seen is it's always the same players you end up with on each team because some guys are just going for so much more than the cheat sheet that you'll never get them.

[DMD]

I am going to look at them again but I can tell you from experience that I have, honestly, received two emails in the same day where one ripped me about how they were far too low in their league and they missed out on all the best players and another said the values were too high and that he overpaid. I am ecstatic that anyone is talking about auctions since they are such a better way to place players on rosters.

[DMD]

What I want to do in the future is to have a way you can make a budget and apply that to rankings. That would be the most accurate way of assigning values to a specific league.

 

I've also had people complain that the auction cheatsheet should only contain the same number of players that are going to be in their league - which I understand in theory - but then you would not be able to draft anyone else other than the exact number applied to the rankings and there could be a very valid reason why you would want a guy ranked 241st when your league uses 240.

 

Again - all great discussions to have. I wish more people would head in this direction.

[Bring Back Pat!!!]

Is it that the tier's reflect the huddle's projection for what players will score, so that is why the projection numbers detlef mentioned are similar, while the % is what you can expect to pay based on actual drafts? Meaning if two guys were similarly projected and one has been selling at 11% and another has been selling at 5%, the lower priced player would be the guy to target?

 

One portion of the cheat sheet is based on projections, while the other is based on draft results, the same way as ADP. ADP will not always fall exactly the order of a cheat sheet's projection.

[SecondString]

"......received two emails in the same day where one ripped me about how they were far too low in their league and they missed out on all the best players and another said the values were too high and that he overpaid."

 

Will happen in every single auction, no two auctions are alike. To expect otherwise would be to expect that you could predict every selling price within a tight pecentage, which is of course almost impossible. Whereas in a snake, the top players will always be near the first selected, not necessarily so in a auction, and nomination order, individual biases, and middle to late draft roster needs have huge effects on bidding prices. Anyone using this tool, their own, or anyone else's AAV's should undestand this going in.

 

Example: Nominate Big Ben real early in a draft, he probably goes pretty cheap, say maybe $11. Nominate him late when there are two owners without a QB yet, and the next best QB left is Hasselbeck, Ben all of sudden becomes a $23 QB, more than doubling his value...no way to predict that.

Edited August 14, 2012 by SecondString

[detlef]

I am going to look at them again but I can tell you from experience that I have, honestly, received two emails in the same day where one ripped me about how they were far too low in their league and they missed out on all the best players and another said the values were too high and that he overpaid. I am ecstatic that anyone is talking about auctions since they are such a better way to place players on rosters.

 

 

Those two aren't mutually exclusive, nor do they point to an unsolvable problem. There are, with out a doubt, players that the Huddle touts in auction cheat sheets at values much, much higher than the market seems to be bearing. And, at the same time, there are guys that you simply will not get if you're only prepared to pay what the huddle says.

 

Mind you, I don't see how that is any different than the huddle saying that AJ Green is worth "spending" a pretty high pick on while another WR may be ranked so low that, provided you're following the huddle's suggestions, you're never going to end up with them based on their ADP.

 

I don't see why auctions would need to be any different.

 

But this is what we pay for. For you to tell us what you think.

 

I know personally, in terms of non-auction cheat sheets, I use them as follows. I'll see you have some guy ranked way higher than I might. I don't throw my list out and go with that guy high, but I at least take a much longer look at them than I would otherwise.

What I want to do in the future is to have a way you can make a budget and apply that to rankings. That would be the most accurate way of assigning values to a specific league.

 

I've also had people complain that the auction cheatsheet should only contain the same number of players that are going to be in their league - which I understand in theory - but then you would not be able to draft anyone else other than the exact number applied to the rankings and there could be a very valid reason why you would want a guy ranked 241st when your league uses 240.

 

Again - all great discussions to have. I wish more people would head in this direction.

 

This, on the other hand, is just a stupid complaint and could be applied to any cheat sheet. "Why are you ranking the 80th best WR? We never get past #50."

 

"Um, because there are bigger leagues?"

[detlef]

After all this and the one league that I've thus-far managed to convert to auction had to go back to snake because of scheduling issues. :kicks rock: