Followers 3

# Apparent bug in statistics

## Recommended Posts

Posted (edited)

For various reasons I'm on a slump at the moment.   To add to the sadness, this seems to have run into a bit of a bug in the statistics.  Have a look at the attached image.

My last find was on June 9, and I'm writing this on July 12. So the slump is from June 10 through July 12, which is 33 days if you include today (21 in June, 12 in July).

So the "Longest slump" line appears to be correct. But the "Current slump" is wrong in two ways. First, the start date of the slump should be June 10 and not June 9. (OK, so maybe the word "since" means something different on that line.)  Second, the length of the slump is *shorter* at 32 days, even though /a/ it's the same slump as the "Longest slump" (so it should have the same length), and /b/ if the calculation Is that it started a day *earlier* than the "longest slump", it should be one day *longer*.

PS: I hope to find a few caches next week...

Edited by sTeamTraen

Hm.

If "Current slump" does not include the 'since' date (per "consecutive days without a find" non-inclusive, and today still incomplete so not included),  and "Longest Slump" is inclusive of the start date and current date (known currently as the period of dates containing no finds, with today 'so far', so inclusive), then the current slump would mathematically be 1 fewer days.

eg, if Current consecutive days slump showed July 9 - July 12 with non-inclusive length 2, then Longest consecutive day slump would show July 10 - July 12 at inclusive length 3.

Yeah, that's confusing if it's how the ranges are calculated. What's inclusive and what's not? Maybe they should be clarified in the text.

Cute bug! Thanks for bringing it up. In my opinion, the problem reduces to one mistake: for the longest slump, today is included in the range, and hence in the count. Since the day's not over, that's a mistake: the slump shouldn't include 7/12 until you get through the entire day without finding a cache. Until that happens, it should say the longest slump is from 6/10 to 7/11, 32 days. With that correction, the two are consistent. Since the current slump obviously raises a question about today, someone took that into account for the current slump code, but apparently they overlooked the special case when the longest slump to notice that it's also the current slump and, therefore, needs the same adjustment. (An alternative would be to include the current date in both counts instead of just the longest count, but I find that less logical.)

I concede that it's a little confusing that the current slump gives the date of the last find while the longest slump gives the first day the slump started, but I like that decision. The longest slump is one thing that can be pointed to precisely, so the dates of the actual slump -- i.e., the range of dates during with you didn't find a cache -- seems like the right thing to report. The current slump is just telling you how long since you found a cache. Somehow it makes sense to me to report the date of the last find, using "since" for current as opposed the first day of the slump and "from" for longest. Any other time, that difference makes perfect sense, so I'm willing for it to be a little confusing when the two lines are talking about the same slump.

Yep, I think the word could just be made a little more clear, because at a glance you see 'slump' and the two lengths, and wonder why (if the current is also the longest) the numbers don't sync.   It takes much more thought to reason out how the numbers aren't incorrect, just using slightly different wordings that could easily be misunderstood.

## Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

×   Pasted as rich text.   Paste as plain text instead

Only 75 emoji are allowed.