Some Boring Numbers...

The only reasonable way to make a projection of how things are progressing while the Herring Run is active would be to look at prior years to see what "percent complete" those years were on the same relative day of the run.

So, for example, on 04-23-2009 we were in Day 5 of the run.

Below shows what percent complete we were on Day 5 in prior years. This is computed by dividing the fish counted through Day 5 by what was counted in total for that year. In the brackets is the projection of this year's count to date [through Day 5] using the percent done for each of those years:

• 2008 - 15% [2,734]
• 2007 - 31% [1,303]
• 2006 - 37% [1,092]

This shows quite a range, which demonstrates how wrong you can be when you have such little history to base projections upon.

A more reliable predictor may be to use a composite of all the prior years. This would be done by adding together the final total of all Herring counted in the prior years and dividing that into the sum of just [in this example] the first 5 days of all prior years.

When we do that [1,584 / 7,693] the resultant "Composite Percent Done" is 20.6%. With this year's cumulative count through Day 5 of 404, that projects out to 1,962 Herring to be counted this year [counted, not total Herring].

This 1,962 is 1,238 more than the average of all prior years counted though Day 5, or 37% better than that average. Note that these numbers will most likely fluctuate as this year's run progresses.

This approach is based upon precious little history [just 3 years now], but may prove to be more useful as we accumulate more data. There may be more precise methods as well with more data. And each year doesn't unfold like the prior year's either, with many factors affecting when and in what numbers the Herring arrive.

I do all of this calculating to answer the frequently asked question [which now seems solely directed at me]: "How's the run doing?" So, now I can say, for example, "Through Day 5, 37% better than the average of the last 3 years."

You can follow these numbers being calculated as the run progresses by clicking on the "schedules and counts" link at the right, then clicking on the "MillPond.YearToYear" tab.

Coming next year: Projecting what the total number that the Herring migration will be, not just the counts.