Jump to content

Enhanced Stats (continued)


Shandon

Recommended Posts

...

So to sort of sum up the responses I've seen on this issue:

...

What did I miss?

Well, here's two more:

 

Jeremy has alot on his plate with planned enhancements to this site. I imagine that including a feature that is not supported by many members is not high on his hit parade.

 

Stats are not the core competency of GC.com. The core feature is the listing of caches. Enhancements to the site should be in line with this core function.

 

A wise man who has considerable pull at GC.com stated that the suggested stats would not be fair and could be easily cheated. Several examples have been given of how any stats would not be truly representative. One thing that I learned in college and has been shown to be true is that if the statistics are known to be incorrect, you are much better without them. They no longer have purpose.

Link to comment

Ah, but there's the rub. Stats can't be unfair if they are not compared to another's stats.

 

There is a difference between a leaderboard and looking at one person's numbers at different angles.

 

There is a difference between my rank in the world and knowing the aggregate distance from home of the caches I've found. Or a visualization of the D/T of the caches I've found. Or a whole host of other angles.

Link to comment
Jeremy has alot on his plate with planned enhancements to this site. I imagine that including a feature that is not supported by many members is not high on his hit parade.

 

Stats are not the core competency of GC.com. The core feature is the listing of caches. Enhancements to the site should be in line with this core function.

 

A wise man who has considerable pull at GC.com stated that the suggested stats would not be fair and could be easily cheated. Several examples have been given of how any stats would not be truly representative. One thing that I learned in college and has been shown to be true is that if the statistics are known to be incorrect, you are much better without them. They no longer have purpose.

What? Because Projects X, Y, or Z are in progress...we should not add Project D to the queue? What sort of faulty logic is that?

 

The "support" for or against statistics is unmeasured (certainly you don't take this thread as some sort of reasonable population sample, I hope), so I'm unsure how you weighed the project's worthiness to be added as a planned project in your statement. Does someone who doesn't care about the site having statistics "support" it or not? I would propose that the level of support against having statistics at this site is actually very low. The majority would be found to be either unconcerned or interested in seeing what may come of it. There are a somewhat even amount of people who would rather see the website shut down first and those who would rather leave if statistics would be completely banned. Given that an opt-out system would satisfy the fraction of people who would see the site shutdown first, I don't see any group who justifably oppose statistics being added to this site in some form at some time. This goes to your second point.

 

The site is a listing service, but there are forums for discussion (not a *core* component). In fact, when the forums went down, they were suddenly a priority replacement because they are a *core* component of what people use here even though they have no application to the ability of the site to be a listing service. This site is primarily a listing service but it's by no means limited to only pointing out the closest cache for you to find. Initially it was to list geocaches and now it has benchmarks. Anything to enhance the end user's interest in the game should be considered an enhancement to this site. Google is a search engine...but now it's even a rudimentary calculator and an excellent conversion tool. The development of a website can easily reach well beyond its "core" components.

 

As for the validity of any specific statistics, it is clearly dependent on what you wish to examine. If I simply want to know what my find to hide ratio is, then I can calculate that easily enough even with Jeremy's rudimentary graphic creator. The statistic is highly valid for displaying how often I find vs. hide caches since I joined. The statistic that Jeremy mentioned as invalid is the combination of difficulty and terrain with your find count. I suggest that he's actually wrong here. While a few mislisted caches may seem to be able to reduce the statistic's validity, over the 50-100 caches the average user might have, these fluctuations will be come insignificant. This is also by no means some sort of "all defining statistic" since that's a fairly self-defeating concept. Baseball (one of the most statistically analyzed sports) has On-Base-Percentage, how often do you get to at-least 1st base, and Slugging Percentage, how far around the bases do you get on your average hit. Nobody looks to slugging percentage as the most important statistic, because it isn't indicative of the guy whose shrewd eye gets him a walk almost every at-bat and he is then a useful base runner and scores runs.

 

None of the statistics I've seen proposed here would actually be invalid at showing exactly what they are meant to show. It is the misinterpretation of these statistics which becomes unfair or unrepresentative. I know that's a fairly lofty concept but it is the truth about statistics. There are very easily ways of looking at the numbers which would provide some very interesting answers (as CR has pointed out) that don't even have to do with "competition" and yet even in competition, some statistics can be calculated that would provide interesting challenges to those that would want to use them that way (FTF per find ratio...etc).

Link to comment

OK, now I want to get back into the fray, having started all of this with the original post.

 

Agree that from the standpoint of 'competition' it would be unfair to calculate scores in the form of (difficulty+terrain) * number found. That would be a problem for persons not physically able to do long hikes, climb rock formations, etc.

 

However, it still would be of interest to see this on my own page, just to have more information on my record -- and if several of us here in south Florida wanted to 'compete' we could agree to do so, using those stats. By the way, when this was done on the Florida site, it was a very popular feature.

 

Anyway, the issue does remain about persons who do not properly rate their caches. I have found some 4-star caches that should have been 1-star, and some 1-stars that should have been 3-stars. However, with nearly 200 caches, it is clear that those are exceptions, not typical. This point was made earlier.

 

One also could overcome this other issue of caches not having the correct difficulty / terrain level by having a simple form on GC.com that is used when posting a new cache -- one could check boxes for various attributes, and the difficulty level could be calculated automatically. I'd have no problem with that, but suspect this idea could open up another can of worms.

Link to comment
Agree that from the standpoint of 'competition' it would be unfair to calculate scores in the form of (difficulty+terrain) * number found. That would be a problem for persons not physically able to do long hikes, climb rock formations, etc.

Someone made this point in a similar thread in the General forum and in response I put forth:

 

So what?

 

Let's call this, [(T+D)*F], the Slugging Percentage of GC.com. A wheelchair-bound cacher will have a lower SLG compared to a mountain climber in their region. So what? That's to be expected...in fact, if a wheelchair-bound cacher had a higher SLG than that mountain climber, it'd be an interesting statistic. But if that wheelchair-bound cacher had a higher completion percentage of 1/1 caches than the mountain climber, then that'd be another interesting statistic and through all of it, we'd see that the wheelchair cacher has been out to most of the easiest access caches while the mountain climber has been concerned with high difficulty/terrain caches...and so on.

 

No one who is interested in statistics should be looking for some sort of *key* statistic that somehow can define a rank order for everything about every cacher's desires or abilities. That's *never* going to appear because of the inherent assumptions and filtering that any good statistic will require. Nobody looks at 1 statistic and defines every baseball player's rank/ability based on it. There are steal kings, 5x5 well-rounded studs, HR kings, and so on. Even in RBIs, you have a statistic that's VERY heavily team- and batting order-dependent...but it's still something to judge a player by for run production and determining who will help you win games.

 

I really think people need to realize that any leaderboard for geocaching statistics will not be like a golfing board where a single rank is established by a single score but instead will have a number of differently calculated statistics based on various factors that will allow you to examine how your numbers fit against others. Just look at any baseball statistics page for an example.

Link to comment

From my standpoint, I'd rather that there were simple(ish) URL's I could submit that would return data about another cacher (with their explicit permission of course) that would eliminate the adornment html and such, and simply list out the GC#, Lat/Long, Date found, Diff/Terr, and a few other goodies in a format easy to parse. I'm quite happy harvesting the information and using my own database storage and my own CPU for computing stats. I would just like a mechanism for extracting the data that doesn't have a huge cost on either the PQ server, or the web server.

 

I don't care whether any particular person wants a stats page or doesn't want one. I know that I have some users that do want to have a stats page, and I'm happy to code it up. I just don't want to execute some kind of long-running query that will tax the system.

 

Then, if you don't want to see your stats or have 'em computed - well, they won't be. Full stop. But, I can satisfy the users that do.

Link to comment
Guest
This topic is now closed to further replies.
×
×
  • Create New...