Category Scarcity

Eugene Freedman adds that there's another reason you might keep a player at more than even his inflated price.

Sometimes you have to consider microvalues when you decide to pay more than par value for players. This really only applies to a few categories (those that are highly concentrated like saves and steals). If almost all of the SB are projected to be kept and you've got a top SB player above par, he may be worth keeping if it makes you competitive in SB, whereas without him the other steals on your roster become valueless.

I call this category scarcity. But the idea is the same: a paucity of a category in the auction (usually steals or saves, as Eugene points out) can lead to prices that skew well past inflation in the auction and don't jibe with the across-the-board league inflation.

Eugene's calculus is correct, and an important piece of the puzzle. However, you may also be better off dumping a scarce category altogether, particularly if you're going to wind up allocating too many of your auction dollars toward steals to the point where you won't be competitive in the other seven categories.

This is what my "regular" American League looked like last year in early March.

Team	Salary	Value	$ to Spend	Auction Value	Total $
Team A	$125	$196	$135	$111	$307
Team B	$113	$172	$147	$121	$293
Team C	$143	$196	$117	$96	$292
Team D	$112	$145	$148	$122	$267
Team E	$117	$146	$143	$118	$264
Team F	$74	$105	$186	$153	$258
Team G	$126	$144	$134	$110	$254
Team H	$100	$115	$160	$132	$247
Team I	$85	$102	$175	$144	$246
Team J	$155	$147	$105	$86	$233
Team K	$83	$85	$177	$145	$230
Team L	$68	$72	$192	$158	$230
Totals	$1,301	$1,625	$1,819	$1,496	$3,121

The salary column was the salaries of the projected freezes in my league. The value was the projected value. The money to spend was $260 minus salary, while the auction value took the remaining dollars each team had to spend and applied the inflation rate to those dollars. A more detailed explanation of this calculation can be found by following the link above.

If you're Team A, B, or C, keeping a somewhat overpriced Carl Crawford or Ichiro Suzuki might not have been a bad idea...depending on your other freezes, the dump chips you had on your roster, and the categories where you were strong or weak. Personally, if I'm Team B or C, I don't like giving up even $1 of value to keep a Crawford or Suzuki and would rather dump or half-dump the category. But, again, I might have decided otherwise along the lines of Eugene's analysis.

However, if you're any team from Team D down to Team L, surrendering any value on your roster is almost definitely a bad idea.

That's because the extra points that your overpriced SB guy is generating will come at the expense of at least one other category, and usually will come at the expense of multiple categories. If you chased Brian Roberts to $33 last year and had to settle for Casey Kotchman instead of grabbing Justin Morneau, it cost you points/value across the board.

Category scarcity also leads to wider disparities across the category earlier in the season, depressing the value of steals or saves in trade. If one owner isn't playing for steals, it depresses the in-season value in trade a little. But if two or three owners decide to bag steals, then the value of those steals is even further depressed.

Now you've not only cost yourself the opportunity to buy better stats in other categories across the board at auction, but you've also overpaid for a commodity that has lost trade value in-season because other owners have recognized that overpaying for steals wasn't the best way to optimize their auction dollars.

Keep all of this in mind when deciding to overpay for one category. It can be a good idea. But, in my experience, it generally isn't unless the conditions are absolutely optimal.