There are a lot of old saws along the lines of "Don't sweat the small stuff." Equally, you will find "Take care of the minutes and the hours will take care of themselves." Comes down to exactly the opposite in meanings.
In my world, both as a bridge player and as a computer programmer, I say take care of the small stuff AND the big stuff, and leave the big middle stuff completely alone.
In bridge, at a tournament pairs event, you'll play 24 hands in a round, give or take a few. Most times, the scores on each hand will range from a big fat zero (of which I am sadly familiar with) to a big shiny 12 (rare for me, but it happens). After the round, it is usual for my partner and I to repair to the local fast food emporium (McDonald's mostly) with a copy of the hand records and our scorecards. Our rule is simple. If we scored three or less, or nine or more, we take a look at the hand. Otherwise, we ignore it.
Why?
Although the almost normal scores of 3.5 to 8.5 could be hiding a gigantic, unusual situation, truth be told, almost all the scores are just that, normal. They deviate from the norm of six, because people at other tables playing the same hands, did something goofy. To spend time on those hands is wasted time. And there just isn't enough time between sessions to eat, formulate evening plans AND waste time. So, if it isn't an anomalous score, we ignore it.
What we look to do is to find ways to do less of what got us a bad score and more of what got us good scores. Sometimes it's as simple as one of us (me) is being too aggressive. Most good partnerships are built on a shooter (aggressive player) and a rock (conservative player). There are the odd exceptions, like Eric Rodwell and Jeff Meckstroth, who are both hyper-aggressive, but then they are two of the greatest card-players of all time and get themselves out of bidding messes. At other times, I've been found guilty of being TOO conservative and a dearth of GOOD scores is directly attributable to my being too meek. That's what these anomaly analysis sessions are all about.
Anomaly analysis can spread to other parts of your life too. For me, that other part of my life is programming ... or rather the results of my programming. I have one client who does a major inventory once a year and monthly mini-inventories. Let me concentrate on the yearly inventory.
When it's done, I do a discrepancy report, which shows the counted stock figures versus what the computer program considered to be in stock at the time of the inventory. This particular company manufactures and sells LOTS of widgets, with an active inventory of some 4,000 part numbers. I'd like to think we have an exact count on MOST of the parts, but I've been wrong ... 15 years in a row.
This year's discrepancy report ran 19 pages (about 1,200 items in all). That's down from 29 pages two years ago and 22 last year. These improvements are directly due to the Floor Supervisor that has supervised the last two countings. But it bugs me a LOT to see the endless sheets of paper pouring out of the printer with these 'mistakes' on them.
Then we take a look at the results and suddenly the disaster tends to lose focus. With parts where it's common to move 10,000 at a time, being out by one or two merits a 'blahhh' response. Well within the degree of error. But not all of the discrepancies merit a blase response. Some of the counts do have a troubling difference. In that case, a recount is called for. MOST times, I'm proven right within a respectable difference.
Some times, not.
We ran an anomaly analysis on the differences each year. Because we kept looking at it, year after year, we discovered a pattern to which parts ended up being wrong ... in BOTH directions of difference. A critical flaw in my software was discovered last year. It's taken a LOOOONG time to find the flaw in my software and I predict another two pages will be cut out of next year's report.
Whether it's sales analysis or individual performance on the basketball court for the Toronto Raptors, you HAVE to maximize analysis time and get back out there doing what you are supposed to be doing. Glossing over the mundane middle isn't a bad strategy at all, if it allows you to maximize analysis time and glean intelligence from all the noise that surrounds such endeavours.
By ignoring the forest, sometimes you can actually find the tree.
No comments:
Post a Comment