An Email From My Wife
Date: Fri, 8 Apr 2011 19:37:27 -0400 (EDT)
Subject: A Special Report from the Castle N Laundry Commission
The following is an analysis of the t-shirt wearing habits of Kyle N., as based on laundry data collected since the beginning of calendar year 2011. It covers 19 loads of laundry, and includes all t-shirt content during that period except one load of laundry from mid-February (omitted because it was washed and folded by the subject himself, offering the commission's investigator no chance to record the relevant data).
During the period under study, there were 91 distinct t-shirt wearing events, as documented by the number of shirts that appeared in the laundry basket. However, those events were distributed across only 30 individual shirts, most of them seeing wear multiple times. The distribution is as follows:
7 shirts were worn once
6 shirts were worn twice
4 shirts were worn three times
8 shirts were worn four times
4 shirts were worn five times
1 shirt was worn eight times
A full list of these shirts may be found in Appendix A of the report.
The mean number of wearings for a given shirt was 3.03. The median was 3.5. The mode was 4. A census of the dresser drawer reveals that the subject possesses 98 t-shirts in three drawers; 68 of these were not worn at all during the period under study, which amounts to 69% of the total.
Breaking the total shirt collection down by color, we find the following categories:
white  -- 28 shirts
black  -- 27 shirts
blue -- 18 shirts
grey  -- 10 shirts
red -- 6 shirts
green -- 3 shirts
brown -- 3 shirts
cream  -- 1 shirt
orange -- 1 shirt
multicolored -- 1 shirt
 A large number of the white shirts were sufficiently discolored as to make it difficult to separate them from shirts which are cream-colored. For the purposes of this study, the only shirt counted as "cream" is the alphabet shirt, a recent purchase whose color is not in doubt.
 Similarly, many of the black shirts have faded to a sufficient degree that they may arguably be called grey. Two borderline cases were placed into the latter category; otherwise the investigator gave semi-black shirts the benefit of the doubt.
White shirts returned the most extreme statistics: they comprise the largest category in the drawers, but only one was worn during the period of the study, on a single occasion. White shirts therefore comprise 3.3% of t-shirts worn, and 1.1% of t-shirt wearing events; furthermore, only 3.6% of the white t-shirt collection saw use during this period.
The statistics for black shirts are potentially complicated by the subject's ownership of two identical black Strange Horizons shirts, which cannot be distinguished by any practical means. However, the commission found that the two were regularly found in conjunction, both appearing in the laundry basket during any given period, so this analysis assumes an equal distribution of wearing across the two.
Counting the identical pair as two separate items, 12 black shirts saw use during this period, which is 44.4% of the total black shirt collection. They made up 40% of t-shirts worn, and 42.9% of t-shirt wearing events.
Taking colored shirts as a single grouping, 16 saw wear, comprising 37.2% of the total colored-shirt collection; they constituted 53% of t-shirts worn, and 47.3% of t-shirt wearing events. Breaking them down into individual categories, the numbers are thus:
blue: 6 worn; 33.3% of blue shirt collection; 20% of shirts worn; 30% of wearing events
grey: 4 worn; 40% of grey shirt collection; 13.3% of shirts worn; 8.8% of wearing events
red: 2 worn; 33.3% of red shirt collection; 6.7% of shirts worn; 6.6% of wearing events
brown: 2 worn; 66.6% of brown shirt collection; 6.7% of shirts worn; 9.9% of wearing events
green: 2 worn; 66.6% of green shirt collection; 6.7% of shirts worn; 4.4% of wearing events
cream: 1 worn; 100% of cream shirt collection; 3.3% of shirts worn; 4.4% of wearing events
It should be noted that the most popular shirt in the subject's collection is blue: the Oshiro dojo shirt, which saw eight distinct uses during this period.
The data strongly supports an "accessibility" model of t-shirt selection: the specimens worn most often are those in the top layer or two of a drawer. T-shirts placed deeper in the drawer are only worn when those from superposed layers are in the laundry basket, making laundry frequency the biggest determinant of t-shirt-wearing diversity. This is a self-reinforcing cycle, as the shirts which have been worn will be returned to the top of a drawer, where they are almost certain to be chosen again before long. Two factors disrupted this pattern: first, the purchase of new shirts during the study introduced "false variety," broadening the number of distinct items worn without involving a larger percentage of the existing collection. Second, in late March or early April the subject folded a stack of black shirts, but left them on top of the dresser rather than placing them in a drawer. This caused two previously unworn shirts and one that had seen only a single incident of use to be introduced into the rotation. Otherwise, the only factor likely to produce much "churn" seems to be the delay of laundry, which is undesirable for other reasons.
The commission firmly recommends a substantial reduction in the t-shirt population. The white shirts, in particular, make up almost an entire drawer which sees only the most negligible use. Based on the evidence, the subject could get by with only thirty t-shirts; however, the commission recognizes the undesirability of such a large reduction in the eyes of the subject. It therefore recommends fifty to sixty t-shirts, the number having been chosen on the following basis: such a collection will fit into two dresser drawers (freeing one up for other kinds of clothing and thereby simplifying the broader clothing-storage situation in the bedroom), while still leaving room for new acquisitions, thus forestalling the need for further cuts in the near future.
The criteria by which the selection is to be made are left up to the discretion of the subject; however, the commission will make bold to offer the following suggestions. Based on the subject's personality, it seems most advisable to concentrate t-shirt retention in the following areas:
1) Shirts with significant personal meaning (e.g. Boston Red Sox World Series)
2) Shirts with humor and/or conversational value (e.g. those acquired from Woot)
Reductions might be focused on other categories:
1) Shirts which have seen significant wear and tear (e.g. holes, fraying collars, etc)
2) Shirts with generic personal meaning (e.g. random conference swag)
3) Duplicate shirts
The actual approach to reduction, however, is left up to the discretion of the subject. The commission welcomes discussion of the topic, and hopes that the assembled data proves useful.
APPENDIX -- Wearing incidents by individual shirt
Bubba Ho-Tep -- 3
Jim's Big Ego -- 3
Penguin -- 4
EFF -- 4
Strange Horizons (2) -- 10
LISA -- 5
Watchmen -- 4
Tori Amos -- 2
Pi -- 2
Dublin -- 1
Freedom -- 1
Respect -- 1
ICFA -- 4
Akamai -- 2
dojo -- 8
Jaws -- 3
Terminus -- 2
raven -- 1
Delorean -- 4
future -- 2
Batman -- 1
Mesoamericans -- 4
Communist Party --2
huckleberry -- 5
Usenix -- 4
alphabet -- 4
dojo -- 1