Auction Draft Prep - Custom Excel

[FantasyHockeyU]

Just curious how some of you go about creating pre draft values for auction drafts. I use the Dobber projections as a starting point but that really is only part of the puzzle.

I am working on a spreadsheet that uses the following logic...
- creating a weighted value for each player using my leagues scoring categories (ex - G(1.5)+A(1)+PIM(.2)+SOG(.1))
- last player picked at each position and his value (smaller spread between best and last decreases draft value)
- finding output per position and allocating a % of my budget to each position (totalling output of all position and finding expected output)
- using this budget with the players relative value to create a "fair" auction value

Anyone else do this type of analysis?

[frankiejay]

I've done something very similar to this with baseball. The math is a great starting point, but it is important not to get too legalistic with it because your best values will be somewhat dependent on how the rest of the league is drafting. I mean this especially as it pertains to positions, because one position may be overvalued by the rest of the league, and at some point you're just going to have to bite the bullet and overspend at that position. You can have the perfect math and great projections but still not draft well if you continue to pass on players and end up with a low-risk, low-reward kind of team.

[FantasyHockeyU]

Agreed, I am trying to do it more or less to identify players with under valued by the Yahoo defaults. I think many take the defaults to heart and that is a big error, the value of center for example changes greatly in a 6 team league vs a 12 team leagues.

[junyab]

you should use fantasy hockey geek's site...it doesn't cost much and does the weight algorhythm/math for you

[Pengwin7]

Auction drafts are all about supply/demand & value created by KeeperContractPrices.

In an inaugural $200, 12-team draft... Ovechkin might go for $48.
In 2nd year of a $200, 12-team draft where teams keep 5 players, everybody is keeping their best "value savings" on a player's worth.
And a re-bid on Ovechkin... he might go for $60.

I like numbers as much as the next guy.
But auction bidding comes from experience on how your fellow GMs will act.

The value of any individual player can fluctuate HIGHLY depending on the front-end winning bids in any draft.


Here's the strategy I've used to success when I used to do Auction Baseball.
HIGH-LOW Strategy.
Take your total budget and divide players into "Groups" of players. You want to set an "equal dollar group" for a given quantity (2, 3, 4 player group, whatever.)
The amount you assign to any given group MUST be higher than the bid on the TOP bid player.

For example, if you know Crosby would go for $57... then you may have to create dividing three "pies" of $66 to total your $200 budget. You can even include your own keepers in any group.
Here, if your budget is split into three pie pieces, so does your roster.
If you roster 18 players, then you work with this:

6 players = $66 budget.

And then from there you THINK about how you make up those groups.
If you want Crosby at $57, then you know you will compliment him with 5 players to which you assign $9. Maybe one $5 player and four $1 players.
Or, instead, you may consider that you'd rather target all 6 players at $11 each.

1. But the value in any "spreadsheet" starts from making a very educated assumption on the high-end bids for the top players.
2. You offset these by determining how many minimum bid players will be drafted ($1) to compliment the high-end bidding.
3. Then you fill in the middle, but even this is NOT SO EASY.

For example, let's consider a 2-team league (for simplicity) where each team has $40 each for 4 players, 8 clear-cut guys in play. ($10/player will be average bid)
These guys are somewhat separated in overall value.

Are the players worth: $14, $13, $12, $11, $9, $8, $7, $6.
Or... are the players worth: $20, $16, $12, $10, $10, $8, $3, $1
???
Since auction players do not go off the board in order of value, it's almost impossible to accurately track and manage.



My long way of saying... put some time on the spreadsheet, but realize that (like a goalie-run in a snake-draft), you really aren't going to accurately peg down a player's draft board demand.

[FantasyHockeyU]

Great reply....

I like the idea of creating groups you presented.

I think the big point in creating this document is not to create perfect values but rather to create better values than the default yahoo rankings and values that many of many opponents will be using. For example when I see Keith valued by yahoo for $14 (estimated), I can likely no he is not going target of mine, back off and get a cheaper, alternative elsewhere for $3 and put that $11 to paying up for one of the few elite goalies (similar to why I love Buster Posey in fantasy baseball).

What we are striving for is output relative to peers...... For simplicity lets just use points as the only variable, a forward who scores 60 points, when the last forward drafted is pegged for 50 is often drafted at a higher cost than a 43 pt D (with the last D drafted pegged for 30). When really the 43 point D adds more relative value to your team. I think many of us understand this concept, but often it is not implemented in a draft.

[frankiejay]

I think I have to disagree with Pengwin somewhat. You CAN use a system that's purely math, but it requires that you make your own projections and then, given a baseline for a replacement player, the formulas should work out what the ideal top-end costs should be. In practice, especially in the first year of a league (or a year to year league) you might as well just start with Pengwin's idea of making an educated guess, because it's easy to fall in love with the mathematical "ideal" value for the players.

This is modestly helpful in year one of a league, but it's particularly valuable in keeper leagues after their first year because then the values can be tailor-made to the adjusted budgets and your own projections. I would run it straight first (just use the formulas) and see what comes out of it, then I would tailor it to your specific situation.

[duducks]

Auction draft I'd try to have a ton of $1 or $2 guys that you aren't attached to, so that you can comfortably play the waivers. I guarantee you, one of your 5 keepers will be a guy that you picked up on FA and get to keep for dirt cheap. A FA turned keeper would cost $2 in your league?

By having the most cheap players, you can either overspend on goalies, or overpay to get 2 first round picks (2 guys in your top 12). I find that a lot of auction managers concentrate too much on the auction cost of a player, and forget about the roster-cost of a player. You still have to fit as many stats as possible into a limited number of roster spots. Bargains are great longterm, but the top loaded team is making it to the finals, and you have to abandon your spreadsheet's logic to get first rounders.

Worst-case scenario, if you have a $75 Malkin, is you have the most valuable rental player available, and get to name your price at the deadline.

[frankiejay]

duducks is completely right, especially considering your league specifics which seem to include a % increase in salary rather than a straight $ amount. That allows you to keep cheapies basically forever, but it should also somewhat inflate prospects' value because there's every reason for teams to try to get guys who might be a couple years away for a pittance because that contract will be good forever. I would add more than normal to high end prospects' value based on your league settings.

[ericdaoust]

This is probably the format that requires the most flexibility from an owner. Even with a nice spreadsheet be prepared to see some wacky things. Just remember that every dollar overspent by your opponent in one place means a player will cost you a dollar less down the line.