FF2010: Keeper League discussion thread

*Legion* wrote:

I've been thinking about whether a fixed percentage inflation is even necessary at all with the qualifying offers, or if it would just be extra unnecessary upkeep.

It seems to me like if nobody is willing to make a higher offer on your player, then he's probably priced right where he's supposed to be.

One thing this scenario seems to miss is that sometimes players get cut because their price tag just gets too high for anyone. So rather than having that $10 player be $10, he's now $11. Do you still keep him around? Probably. What happens when he's a $10 player for $15 or $20? Now you're starting to think about cutting him due to cap space not because he's a bad player but because he's not worth the money he's owed.

How many players got cut this offseason not because they were bad but because they were overpriced for their contribution? (rhetorical)

That does raise the question: Do we want to try and simulate that? And the answer may be no, but it feels like that's what you lose without inflation. Also, doing qualifying offers on every player seems like it might be overkill, but what do I know

The flip-side of my so-called "doomsday" scenario is that someone should be rewarded for taking a chance on a sleeper. Perhaps I'm missing something but doesn't this whole qualifying offer kind of kill the whole "keeper" aspect of it? I mean, it sounds like you basically have to pay market price for all your players every year.

As for the whole "do we want to try to simulate..." question, my personal preference for having an auction/keeper league is trying to approximate (distantly, but still...) the NFL General Manager experience. We all know that most NFL contracts are back-loaded so yes, I think inflation would be a good thing to have.

Grumpicus wrote:

The flip-side of my so-called "doomsday" scenario is that someone should be rewarded for taking a chance on a sleeper. Perhaps I'm missing something but doesn't this whole qualifying offer kind of kill the whole "keeper" aspect of it? I mean, it sounds like you basically have to pay market price for all your players every year.

The part you're missing is that the ability to make qualifying offers with compensation allows me to make my keeper cheaper for me than he is for you.

I have a $15 player. I give him a $30 qualifying offer. I've increased his dollar value by $15, which means compensation for poaching him from me would be $15. He now costs $30 for me, but for you to poach him, it would cost you at least $46 ($31 + $15 of compensation to me). And that's just to beat my offer by $1. I can decide to match your $31 and keep the player. And if I let him go, his $30 is off my books, plus I get an extra $15 to work with. Kind of like compensatory draft picks.

So, the strategy is to make qualifying offers that are too expensive to easily outbid and poach, but still low enough to be a deal.

We have to think about the numbers and kick some scenarios around in our head. The idea of compensation being exactly proportional to the qualifying offer's increase was just an idea of a simple way to do it. I don't necessarily know if it's optimal, or if maybe compensation should be some other percentage of the offer increase (more than 100%? less?) or perhaps even some fixed-value format (less interesting, but maybe there are holes in the proportional compensation idea that we need to uncover).

I absolutely agree with you that players should be rewarded for foresight and landing good young talent ahead of their value explosion. The mechanism to make this happen is the asymmetry of cost that compensation creates. We have to think it through to make sure the proportion is correct.

If you're looking for a fill-in, I'd be interested as well.

kaostheory wrote:

How many players got cut this offseason not because they were bad but because they were overpriced for their contribution? (rhetorical)

That does raise the question: Do we want to try and simulate that? And the answer may be no, but it feels like that's what you lose without inflation. Also, doing qualifying offers on every player seems like it might be overkill, but what do I know :D

I think this is why some sort of percentage based price increase is still necessary- ideally we'd be able to find a balance where players are still being cut because their salaries are higher than what the market will bear, and some other players priced below the market are being brought closer to their real value via qualifying offers.

Grumpicus wrote:

The flip-side of my so-called "doomsday" scenario is that someone should be rewarded for taking a chance on a sleeper. Perhaps I'm missing something but doesn't this whole qualifying offer kind of kill the whole "keeper" aspect of it? I mean, it sounds like you basically have to pay market price for all your players every year.

As for the whole "do we want to try to simulate..." question, my personal preference for having an auction/keeper league is trying to approximate (distantly, but still...) the NFL General Manager experience. We all know that most NFL contracts are back-loaded so yes, I think inflation would be a good thing to have.

I think the biggest part of this that none of us have tried to model yet is that an auction dollar this year isn't necessarily worth the same as it was last year, or as it will be in following years. The hard part to predict, in my opinion, is whether the compensation provided to a team that loses a player they gave a qualifying offer, or if it is too large (unlikely, I think, but we'll see), and whether the year to year swings in auction value inflation will unbalance the whole thing. Of course, the majority of this also applies to flat and scaled percentage auction value increases.

BlackSheep wrote:

Perhaps you don't know how this game works? That's a quaint little dig there, though with all that glorious hyperbole built in. Hell, I could let my dog pick my picks and be as likely to be competitive as if I picked them myself, so there's that.

No offense meant, just a poor attempt at humor.

Landshrk83 wrote:
kaostheory wrote:

How many players got cut this offseason not because they were bad but because they were overpriced for their contribution? (rhetorical)

That does raise the question: Do we want to try and simulate that? And the answer may be no, but it feels like that's what you lose without inflation. Also, doing qualifying offers on every player seems like it might be overkill, but what do I know :D

I think this is why some sort of percentage based price increase is still necessary- ideally we'd be able to find a balance where players are still being cut because their salaries are higher than what the market will bear, and some other players priced below the market are being brought closer to their real value via qualifying offers.

Grumpicus wrote:

The flip-side of my so-called "doomsday" scenario is that someone should be rewarded for taking a chance on a sleeper. Perhaps I'm missing something but doesn't this whole qualifying offer kind of kill the whole "keeper" aspect of it? I mean, it sounds like you basically have to pay market price for all your players every year.

As for the whole "do we want to try to simulate..." question, my personal preference for having an auction/keeper league is trying to approximate (distantly, but still...) the NFL General Manager experience. We all know that most NFL contracts are back-loaded so yes, I think inflation would be a good thing to have.

I think the biggest part of this that none of us have tried to model yet is that an auction dollar this year isn't necessarily worth the same as it was last year, or as it will be in following years. The hard part to predict, in my opinion, is whether the compensation provided to a team that loses a player they gave a qualifying offer, or if it is too large (unlikely, I think, but we'll see), and whether the year to year swings in auction value inflation will unbalance the whole thing. Of course, the majority of this also applies to flat and scaled percentage auction value increases.

BlackSheep wrote:

Perhaps you don't know how this game works? That's a quaint little dig there, though with all that glorious hyperbole built in. Hell, I could let my dog pick my picks and be as likely to be competitive as if I picked them myself, so there's that.

No offense meant, just a poor attempt at humor.

Haha.. none taken. Sometimes message boards don't convey subtlety and tone very well. I didn't mean to be an ass either. Apologies here as well for the tone.

*Legion* I made you co-commish (days ago). If you have a 12th player, send them an invite already.

Also, if you (or any of the other co-commishes) want to go ahead and get the scoring setup started, feel free.

ukickmydog: Check your email for an invite

*Legion* wrote:

ukickmydog: Check your email for an invite

In.

Ok, I'm officially back in town (got in last night) from a two-week vacation and will start grinding away on football stuff shortly. We've got 4 days left to see if there's going to be a third GWJFFL division and then I'll start figuring out draft dates. I'm also ready to see if we can get another episode of Booth Review recorded so y'all start harassing *Legion*, Boogle, and Rat Boy.

DRAFT PROPOSAL

Grumpy and I kicked around some ideas for how to conduct the auction draft. Here is a proposal for you guys to consider and comment on.

Speaking just for myself here, I'm of the opinion that auction draft ought to mean non-blind. I think that's what makes it fun. But the trick becomes managing it. I had assumed some sort of mailing list, or Grumpy's suggestion of a collaborative document. But a mailing list is a lot of overhead (gotta open each individual email to see each bid), and a collaborative document lacks temporality.

So, the proposal is to draft via Twitter.

Why Twitter?

* Twitter is optimal for displaying small bits of data in a timeline. Like fantasy auction bids. It is extremely easy to "skim" and the timeline-based presentation makes the sequence of events very clear

* Twitter is easily accessible from the web as well as mobile phones

How?

* Every post goes to a hashtag, like #gwjdraft. Follow that hashtag, and you get a clear timeline of the draft unfolding. You don't have to review every bid to see what the current high bid is: the high bid is the one at the top!

* Twitter timestamps take care of any issue of near-concurrent bidding. The Timestamp Is Law. Issue solved.

The Procedure

I'm less rock-solid on this, so I'm definitely open to comments or ideas. But here's the model I built in my head on just an initial thinking-through:

1. A player is nominated for bidding. In-person auctions tend to do this in a round-robin format, but I think the easiest way for us would be to have the winning bidder of the previous player make the nomination for the next player and the opening bid.

2. Bidding is open. People begin to place bids. (A time limit would exist in the event that nobody else bids on the player - which will happen when we get to the end of the auction and a bunch of $1 players)

3. When you place a bid, you are "in" the bidding. That means you are considered a participant of the bidding until you say you are "out".

4. Bidding ends when there's only one player left "in". (A time limit would exist in the event of an "in" bidder failing to declare himself out, while also failing to place a bid to compete with the current high bid)

5. Note: If you have taken yourself "out" of the bidding, you are free to jump back "in" if you want (just place a bid, and you're "in" again). But the bidding will end when there is only one player left "in", so you would have to enter that bid before the next-to-last person bows out.

6. Player is awarded. Winner nominates the next player and the cycle begins.

It would look something like this, except the bids would be from your Twitter accounts:

IMAGE(http://a.imageshack.us/img835/4946/exampledraft.jpg)

Very easy to follow and see what's current. (Just a rough example, there'd probably be a post that's actually from gwjfflk declaring the end of bidding on one player and the start of another - but maybe not, maybe bidders can be trusted to recognize the end of bidding and kick off the next one)

I think this is very clear, cuts down on all unnecessary chatter, and doesn't force Grumpy or I to have to set a global time-limit for every single bid. There would just be a set "timeout" period, which closes bidding if a certain length of time has passed from the timestamp of the last bid (but we would greatly prefer if people remember to declare themselves "out" so we don't have to wait through this).

I have also thought of the possibility of augmenting this with email notifications when bidding closes & the next bidding opens, as a reminder. These could even be sent through SMS gateways to people that want to receive them as text messages.

Please read this over and give us your comments.

Oh man. As a first time auction drafter that is very intimidating for me. I am used to the general serpentine draft process that means I only get to make real decisions once every 24 players or so. I'm just realizing how intense it will be to bid / interact multiple times for many players. Over what period of time are we talking about doing this? A couple evenings where we all try to get online at once? A couple weeks of back and forth waiting for responses/time outs?

How much of this is just me being a baby? I'm looking forward to my picks mystifying all of you. I'm prepared to be the Al Davis of GWJFFLK.

So what was the issue with doing an auction draft on a site that has clients for them?

ukickmydog wrote:

So what was the issue with doing an auction draft on a site that has clients for them?

The fact that the ones I've seen are all "live" real-time drafts. It's tough in the regular leagues to get everyone to a live draft. Someone is always missing. But at least those drafts have a sensible auto-drafter. Auto-draft is tough in an auction draft.

Even in the event of finding one that isn't real-time, Grump and I both have the desire to allow drafters the ability to draft from their phones. Everyone is in different time zones and people are on the go, it's nice to be able to drop in a bid from your phone.

Jolly Bill wrote:

Over what period of time are we talking about doing this? A couple evenings where we all try to get online at once? A couple weeks of back and forth waiting for responses/time outs?

Between now and the end of the preseason, I think.

In practice, auction drafting is very top-heavy - the "early round" guys get a highly disproportionate amount of the bid money. After that, it drops off sharply, and the latter part of the auction ends up being very much draft-like: everyone just goes for the $1 opening bid.

This does help me identify a catch in my proposal, though. When we get down to the end, if a guy bids $1 and gets the player, by my proposal he would get to nominate the next guy that's probably only worth $1, and so on. I guess the rule tweak would be that the winning bidder only nominates a player if there were competing bids on the player he won - otherwise the ability to nominate a player moves on to someone else.

I like the idea. I just worry that somebody will get caught up with work for 1-2 days and miss out on drafting Beanie Wells who is their favorite player and they would've gladly paid $30 for but didn't get the chance as they hoped Beanie wouldn't be selected for auction at that time.

Y'know what, I just thought of a possible solution to that scenario. Scheduled auctions. We pick a ranking site, let us say Yahoo's. On Day 1, we have players 1-10. On Day 2, we have players 11-20, day 3 = 21-30...etc

This way I can see at the very beginning that my man Beanie is ranked 28th. Which means I can bid for him on day 3. Once day 3 opens up I can post:

Elliottx: $30 Beanie Wells #28 #gwjdraft (Once again I dunno if we want to do the number for the player or the name or both)

Since it's day 3 everybody knows that Beanie Wells is up for bid then. Now this does raise the worry that bidding for 10 different players each day could cause some confusion but it shouldn't be too hard as I think most players will only get 2-3 bids as there is no point in putting in a bid for Beanie Wells at $5, instead we'll see:

Legion: $21 Beanie Wells #28 #gwjdraft
Elliottx: $30 Beanie Wells #28 #gwjdraft

And no other twits mentioning Beanie Wells or #28. Sifting through 40 twits at midnight EST every night and awarding players to who posted the highest numbers wouldn't take long at all. Heck I'll volunteer to do the sifting and updating of the auction spreadsheet in this scenario.

Elliottx wrote:

I like the idea. I just worry that somebody will get caught up with work for 1-2 days and miss out on drafting Beanie Wells who is their favorite player and they would've gladly paid $30 for but didn't get the chance as they hoped Beanie wouldn't be selected for auction at that time.

Y'know what, I just thought of a possible solution to that scenario. Scheduled auctions. We pick a ranking site, let us say Yahoo's. On Day 1, we have players 1-10. On Day 2, we have players 11-20, day 3 = 21-30...etc

This way I can see at the very beginning that my man Beanie is ranked 28th. Which means I can bid for him on day 3. Once day 3 opens up I can post:

Elliottx: $30 Beanie Wells #28 #gwjdraft (Once again I dunno if we want to do the number for the player or the name or both)

Since it's day 3 everybody knows that Beanie Wells is up for bid then. Now this does raise the worry that bidding for 10 different players each day could cause some confusion but it shouldn't be too hard as I think most players will only get 2-3 bids as there is no point in putting in a bid for Beanie Wells at $5, instead we'll see:

Legion: $21 Beanie Wells #28 #gwjdraft
Elliottx: $30 Beanie Wells #28 #gwjdraft

And no other twits mentioning Beanie Wells or #28. Sifting through 40 twits at midnight EST every night and awarding players to who posted the highest numbers wouldn't take long at all. Heck I'll volunteer to do the sifting and updating of the auction spreadsheet in this scenario.

I've never done auction drafts, but i think this pretty much defeats the purpose of it. Part of the strategy in auction drafts is when to nominate someone. Though what Legion proposed with whoever wins a player, gets to nominate the next player, also kind of ruins that...

Ahh see I've never done auction drafts but I do remember hearing that the nomination of players is part of the fun. I forgot about that.

I'm not sure there is any way to avoid a rotating nomination cycle while still keeping the system fair.

Every new day everybody gets to nominate one player? If people overlap on their nominations, no re-do. Once again I'm pushing the idea of a 24 hour auction window instead of the "out" scenario.

I'm good with Twitter. That seems like an interesting idea.

*Legion* wrote:

1. A player is nominated for bidding. In-person auctions tend to do this in a round-robin format, but I think the easiest way for us would be to have the winning bidder of the previous player make the nomination for the next player and the opening bid.

Why would this be easiest? Is it because we'd be nominating and betting over the phone so keeping track of who's up next is difficult? If so, couldn't the admins facilitate the round robin? So, you and Grumpicus (or whomever) could have the draft order and once bidding on a player is over, you guys could say who is up to pick. Of course we could also post the draft order here and the admins could lightly facilitate the draft, nudging people when they didn't step up and take their turn.

Just email me instructions when this is starting commish. My manboy brain cannot stand this thinking.

ukickmydog wrote:

Part of the strategy in auction drafts is when to nominate someone. Though what Legion proposed with whoever wins a player, gets to nominate the next player, also kind of ruins that...

I would say it's still very much part of the strategy. I don't see how shifting from winning bidder to round-robin changes that - the person with the power to nominate is going to be very deliberate about who they nominate. And everyone needs to win bids to get players, so it's not like nomination is going to get monopolized by one person.

But I am not against round-robin. I just saw winning bidder as a way to keep draft momentum going by having the last person who did something be the one to nominate and keep things moving.

garion333 wrote:

Why would this be easiest? Is it because we'd be nominating and betting over the phone so keeping track of who's up next is difficult? If so, couldn't the admins facilitate the round robin? So, you and Grumpicus (or whomever) could have the draft order and once bidding on a player is over, you guys could say who is up to pick. Of course we could also post the draft order here and the admins could lightly facilitate the draft, nudging people when they didn't step up and take their turn.

I replied to ukickmydog before reading your post - I just sorta answered that in the reply to him. It just prevents any sort of out-of-band management. You know who is next by looking at the screen: the name of whomever just bid last is the guy that needs to offer up the next player for bid. It's a zero-management solution, nobody needs to get nudged and reminded, "hey, it's your turn". It's just plain obvious who does what next.

As you know from this thread, my bias is towards things that are self-managing. I'm not opposed to the alternative, I'm just suggesting this in an effort to be as streamlined as possible.

*Legion* wrote:
garion333 wrote:

Why would this be easiest? Is it because we'd be nominating and betting over the phone so keeping track of who's up next is difficult? If so, couldn't the admins facilitate the round robin? So, you and Grumpicus (or whomever) could have the draft order and once bidding on a player is over, you guys could say who is up to pick. Of course we could also post the draft order here and the admins could lightly facilitate the draft, nudging people when they didn't step up and take their turn.

I replied to ukickmydog before reading your post - I just sorta answered that in the reply to him. It just prevents any sort of out-of-band management. You know who is next by looking at the screen: the name of whomever just bid last is the guy that needs to offer up the next player for bid. It's a zero-management solution, nobody needs to get nudged and reminded, "hey, it's your turn". It's just plain obvious who does what next.

As you know from this thread, my bias is towards things that are self-managing. I'm not opposed to the alternative, I'm just suggesting this in an effort to be as streamlined as possible.

Maybe I'm being a little anal here, but wouldn't it be the person who just lost the last bid who would be the last to post in the twitter account? In the example above, the second "out" post? I would be fine with that as the self-managing auction. Also falls apart logically when we hit the later rounds anyway, since there will be times when no one bids against the nominator, so we could move on to round robin when that starts happening. It requires the last person 'out' to know who they are, but then again the winner nominating also requires someone to know when everyone is out, so I guess that's not a real difference.

Jolly Bill wrote:

Maybe I'm being a little anal here, but wouldn't it be the person who just lost the last bid who would be the last to post in the twitter account? In the example above, the second "out" post?

I suppose, but it makes more sense to me for the winner to be the one that gets to nominate the player. Otherwise, you have the possibility of people hanging around in bids to try and be the last one to bow out, prolonging a bid needlessly.

I figure the winning bidder is more likely to be following a particular bid closely than a random other person, so they'll be more likely to be ready to respond with the next nomination once a bid closes. After all, the winner is watching in case that last "out" was actually a competing bid instead.

Round-robin in the later stages makes sense to me. I guess the formula would be: winning bidder if the bid had other competing bids, or round-robin to the next owner if the last bid went unchallenged.

I like the Twitter idea a lot, especially because it lets me keep up with an extended auction process from work, travel, etc.

However, one thing that worries me a little bit is the time element. Assuming we're using the same size roster as the normal GWJ leagues, we've got 240 players to draft before the season starts on Sept 9. If we started the draft today, that gives us exactly 3 weeks to get things squared away, so we'd need to get about 12 players auctioned off each day to finish on time. If we're conducting the auctions one player at a time, that doesn't leave much time per player, and worse yet it would mean that if anyone had a couple hour period of time that they couldn't watch the auction, they'd probably miss players entirely.

The idea of nominating players with a number of auctions going on each day is interesting, but it does take away some (a lot?) of the strategy of auction drafting. Also, if there were a half dozen or more auctions going on every day, it would get hard to track, even with Twitter. I realize I'm not offering any alternatives here, but I'm not sure there's a really good solution.

A potential compromise (feel free to shoot it full of holes)- nominate players round-robin style, and start doing the nominations a few days before the actual auctions start, the idea being that the nominations stay a few rounds ahead of the actual auction going on. Nominated players could either be assigned time slots for bidding, or have a whole round auctioned at a time, either of which might pose problems.

I think the only sure thing is that we will not finish the draft in time if we auction one player at a time without a pretty strict timeline, either bidding on multiple players at a time, or adhering to strict time slots for bidding.

Maybe we help mitigate the number of people we need to auction draft by not auctioning full rosters. We could auction draft most of our roster and just fill out the rest with pre-season free agents. Beyond this year, we're not going to draft full rosters anyway - we'll all have keepers. We just need to figure out a way to survive this initial "seeding" of our rosters.

No question, we have to get moving in the next day or so.

Another possible idea: auction draft the first X number of players on Twitter like proposed, then get together one night for an online "draft" for the rest - all those $1 guys. Anyone that can't make it could submit to Grump or me a "wish list" of guys to propose/take on their nomination turns. (Since the guys are all likely to be $1 on there, we don't have to worry about managing the absentee's "bidding").

That lets us use the next couple of weeks for all the major auctioning. And it makes the get-together online draft night not the kind of death march it would be if we did the full auction that way.

The floor is open to all ideas.

I like that idea. Just the first 10 rounds as Auction, then the final 10 rounds as snake. Since we're all going to want to hunt for those sleeper players, we really could become muddled on the likes of Early Doucet. Or do the final 10 rounds as blind bids via Free Agency...That way we do 10 rounds of visible auction and then do 10 rounds of blind bidding for sleepers.

Elliottx wrote:

I like that idea. Just the first 10 rounds as Auction, then the final 10 rounds as snake. Since we're all going to want to hunt for those sleeper players, we really could become muddled on the likes of Early Doucet. Or do the final 10 rounds as blind bids via Free Agency...That way we do 10 rounds of visible auction and then do 10 rounds of blind bidding for sleepers.

I'm liking the suggestion, but...

Even at 10 rounds we're talking 120 players...without some sort of timed structure in place it's going to be tough to get that done in time to start the season.

Even with a couple hours set as the default time a player's auction ends without bids, we could easily waste a few weeks getting through the first couple rounds.

Just a thought, but what about setting up a survey at whenisgood to see if there is a time everyone could make a live draft? In the unlikely event that everyone can make a certain draft time then we can get the full auction treatment. If not, then we pursue the Twitter option?