Friday, June 05, 2009

What Can Skew Do For You?

Late in March, Eugene Freedman explained how he "skews" his prices for leagues he's in with inflation:
I skew my inflation dollars toward the top players. I don't distribute it linearly. I delete any inflationary dollars that have been attributed to players $10 and under and redistribute it to players over $25. If I add $1 to everyone and there still is money left it goes to those over $30 and then when that's done over $35 and $40 until it's all gone. If there is a huge amount of inflation I might start at $20.
I tend to "skew" prices based on my own preferences, roster strengths and weaknesses, and other considerations. What if I had skewed simply using Alex Patton's raw prices?

2009 Billy Almon Brown Graduate w/Skew
#Player$*
Bid
Inf
Skew
Cost
1Carl Crawford
$52$35$44
$47$43
2Matt Holliday
$19$30$38
$41$42
3Ichiro Suzuki
$34$29$36
$39$36
4Mark Teixeira$28$29$36
$39$41
5Magglio Ordonez
$8$28$35
$37$32
6
Vlad Guerrero
$4$27$34
$36$36
7
Dustin Pedroia
$27$27$34$36$41
8Alex Rodriguez
$10$27$34
$36$30
9
Curtis Granderson
$27$27
$34
$36$35
10
David Ortiz
-$12$25$31
$33$24
11
Kevin Youkilis
$32$25$31
$33$31
12
Bobby Abreu
$26$23$29
$30$30
13
Johnny Damon
$28$23$29
$30$29
14
Torii Hunter
$39$23$29
$30$29
15
Carlos Gomez
-$1$23$29
$30$26
16
Jason Bay
$34$23$29
$30$34
17
Chone Figgins
$28$22$28
$29$29
18
Vernon Wells
$14$22$28
$29$26

Average
$22$26$33
$35$33
*RT Player Rater: 6/5/09, 4x4, 70/30 hitting

The "cost" column is what these players actually went for in my non-expert American League-only league.

Right away, we can see that some owners "skew" even if that isn't the term to use to define what they are doing. Pedroia, Bay, Teixeira, and Holliday all go higher than even their skew prices. This could be an accident, but I think that my league is insisting that these players are worth a certain amount and should be paid accordingly.

Looking at this group on the whole, I don't think I like the idea of skewing hitter prices using a method that doesn't have manual intervention.

I do not have photographic recall of my pre-auction bids, but I do remember that I had added $2-3 dollars to Bay's raw bid and took a buck or two away from Magglio Ordonez. It's likely I made some other minor tweaks, but those are the two bids that jump out at me.

On the whole, though, I don't like the idea of adding $2 per player to a group of players that loses $4 per player before inflation even kicks in.

One concept that I repeat over and over and over again during the course of writing this blog is that there is a ceiling on what the top players can possibly earn. At some point, your bids are going to go past that ceiling in leagues with moderate or heavy inflation and you're going to overpay the best players. Overpaying the best players by slight amounts is OK; we're still getting a lot of stats from the best players, and a nominal loss is acceptable.

But taking a $6 bath on the top players is unacceptable.

It's apparent that a number of owners are using their own version of the skew on certain players. Mark Teixeira, Jason Bay, Vlad Guerrero...owners are making sure that they buy the best players available even if it costs an extra buck or two. Whether this is the skew at work or not I can't say. Chances are that owners are making adjustments like Eugene is.

I suspect, though, that their adjustments are manual like mine and not simply on the best players in the auction. I added money to Tex and Bay primarily because I wanted power. I needed speed, but decided not to add money to Crawford and Suzuki because I had a weak freeze list and didn't want to allocate too much money toward those players.

As I've said before, the best place to do this is in your pre-inflation bids. If you thought Magglio's $28 pre-inflation bid was too high (like I did), you were better off pushing that number down before you started adding inflation to your bids. It could just be me, but if I saw $28 on my sheet for an un-inflated Mags I might have said $33 at my auction and fooled myself into thinking that I was going to be buying a bargain.

3 comments:

Eugene Freedman said...

If you didn't like Ordonez, why did you project him to be worth what you projected him to be worth. That's really what it comes down to. I make my projections based upon various factors, but then I downgrade players who I don't like in their projections, either in playing time or individual stats. So, if I don't like a particular player his bid is affected, so I don't have to subjectively lower and raise bid values. The subjectivity is in the projections.

Eugene Freedman said...

Just carrying that thought forward - when the bid value is lower, then the inflation is lower and the skew value doesn't come into play.

Mike Gianella said...

Wasn't my projection. I was going off of Patton's projections for the purposes of discussion. My bids were skewed greatly in favor of MI and as a result I wasn't comfortable including these for the purposes of this article.