Jump to content

If you were constructing an algorithm ...


Recommended Posts

24 minutes ago, arisoft said:

 

Good question. It is practically the same I asked earlier, why quick response to NMs or DNFs is more valuable than good caches. My answer is that for me the current situation matters when we are speaking about new (virtual) caches. If we want to honor someone for long career the new (virtual) cache may not be the best or eligible award at all. The honorary membership suit better for that situation.

But then you're effectively rewarding popularity, and not quality (in as much as FPs are a measure of quality) as @funkymunkyzone said earlier.

Consider 2 caches placed by 2 cachers:

1. A high quality cache placed >1 year ago. All the local cachers found it within the first 6 months and gave it a lot of FPs, but since then it only gets infrequent visits from new joiners or people passing through.

 

2. A mediocre cache placed 6 months ago. All the local cachers have found it in the last 6 and it has collected  some FPs.

 

Your suggestion would favour the second, mediocre, cache over the first high quality cache, effectively you're rewarding people for putting out caches frequently,  and discouraging them from maintaining high quality long lasting caches.

 

 

  • Upvote 1
Link to comment
23 minutes ago, MartyBartfast said:

But then you're effectively rewarding popularity, and not quality

 

If you mean the number of finds as popularity, you may be right, but this is not so easy and straightforward problem that you can skip the idea by picking just some unsolved situation as a criterion. As you have seen, every method seems to have some unsolved problems. I am sure that you do not want to award unpopular caches either.


There is no such premise that popularity shouldn't count. Is there something wrong with popularity? Why do we want to award somebody? The explanation is here https://bonus.ly/employee-recognition-guide/types-of-employee-recognition "Organizations across the world see the need to improve engagement"  The algorithm should pick players whose engagement is the most valuable to the business compared to the risk that they are going to leave. From this view giving the award to the worst player may be the obvious strategy. Right?

Link to comment
2 minutes ago, arisoft said:

The algorithm should pick players whose engagement is the most valuable to the business compared to the risk that they are going to leave. From this view giving the award to the worst player may be the obvious strategy. Right?

 

Ah - so you think the algorithm should reward based on how many $$$ a player puts into Groundpeak's coffers.

Link to comment
48 minutes ago, MartyBartfast said:

Your suggestion would favour the second, mediocre, cache over the first high quality cache, effectively you're rewarding people for putting out caches frequently,  and discouraging them from maintaining high quality long lasting caches.

 

Practically yes if no one is visiting the "better" cache. Caches vithout visitors are not the most valueble to the business at any level.

Link to comment
Just now, Team Microdot said:

Ah - so you think the algorithm should reward based on how many $$$ a player puts into Groundpeak's coffers.

 

This is not my opinion it is the way the business runs and you should adapt to this fact just because it is a fact that Groudspeak must run the business for their shareholders.

Link to comment
23 minutes ago, arisoft said:

There is no such premise that popularity shouldn't count. Is there something wrong with popularity?

There is nothing wrong with popularity, but it doesn't really indicate quality either. Unless you're willing to say that McDonalds and KFC are among the best restaurants in the world.

Link to comment
2 hours ago, arisoft said:

Both of these examples are quite old caches and I think that they do not collect so many favorites now.

 

The second one may be older but it accrued a majority of its FPs within the past three months, particularly at the end of May of 2018 when the Giga event, Geowoodstock was across the river in Cincinnati.  You didn't specify that the cache needed to be new, just that the FPs are recent, which the second clearly meets.  You can't keep changing the "rules" you're attempting to establish in order to meet the criteria you're hoping to prove or disprove.

Link to comment
2 hours ago, arisoft said:

My answer is that for me the current situation matters when we are speaking about new (virtual) caches.

 

So that means that new cachers who have placed a cache within the past three months and have received some FPs on a regular basis are more likely to receive a theoretical virtual reward than a cacher who has been placing caches for 5 years or more, receives some FPs on a somewhat regular basis (but not a lot recently because it's not visited very often as all the locals have already found it), has very few, if any maintenance issues, but hasn't put out a cache in the past year.  Seems to me you'd be rewarding the wrong person in the situation above, based on your "current" criteria.

Link to comment
8 minutes ago, coachstahly said:

You didn't specify that the cache needed to be new, just that the FPs are recent, which the second clearly meets.

 

I didn't and it was intentional. So that cache was the best of the event :) and there was no maintenance problems with the CO. Usually large events will cause massive need for maintenance due to full logs and misplaced caches.

Link to comment

You make it sound like making someone laugh has ulterior motives just to get favorite points.  Which I completely disagree with. I speculate that it is just to make people laugh. It's hard to make a film canister funny and besides I bet the kids love it. And honestly we need more things that get kids both older and younger excited.

 

For me the biggest let down was going after the #4 favorited cache in my region https://coord.info/GC12  with 1054 FPs (51% though) it was just a bucket full of water with a cracked lid and the destination was nothing special other than an empty calendar space.  While GC17 a few miles away get the #6 spot with far fewer signers 400+ has 816 FPs (55%) and has on a good day an amazing view at the end of the hike.  

 

I don't think you will ever understand what motivates people to give FPs. Maybe the view was covered with clouds. They had a great group of friends that went to a special area that they enjoyed. Maybe it was poring down rain and really cold. Maybe they laughed at the container. Maybe they were just in a good mood. Or it's a plastic turkey with a bison tube sticking out of that your kid could not stop laughing about.

 

 

 

Link to comment
3 minutes ago, arisoft said:

So that cache was the best of the event

 

Nope.  There were others that received more FPs than this one, particularly as you search closer to the actual event location.  I just singled this one out because you asked for examples of caches with recent high FPs being awarded that aren't "good".  I think it's cute, but certainly not a good cache, as I would define it.

Link to comment
2 minutes ago, MNTA said:

I don't think you will ever understand what motivates people to give FPs.

 

Some COs are motivated to get high number of favorites and they know how to do this. But it is somehow pity that there are also many COs who makes great caches but do not earn favorites because they focuses other features like endurance etc. not the wow effect.

Link to comment
3 minutes ago, coachstahly said:

Nope.  There were others that received more FPs than this one, particularly as you search closer to the actual event location.  I just singled this one out because you asked for examples of caches with recent high FPs being awarded that aren't "good".  I think it's cute, but certainly not a good cache, as I would define it.

 

Ok - so it was not at the winning position at all.  We must remember that the algorithm takes account all caches from the same CO and a single cache do not have great effect. We are not giving awards to caches but their owners. More caches, active maintenance, more visitors, more wow effects all have effect to total number of favorites.

Link to comment
4 minutes ago, arisoft said:

Just count the favorites for some specified period, for example last year.

 

Your argument for using current stats (which has evolved from current to a year long period of time) eliminates all caches that have been out for a few years, received FPs, but don't receive the same number of FPs they did originally because the locals have found them and they only get the occasional visitor now.  I fail to see how that's an equitable measurement for determining whether or not someone is in the running for a possible theoretical virtual reward.

  • Upvote 2
Link to comment
8 minutes ago, arisoft said:

We must remember that the algorithm takes account all caches from the same CO and a single cache do not have great effect.

 

How many people complained about the cachers who received a virtual reward who only had a few caches that had been placed?  In those situations, a single cache can have a HUGE influence so that argument doesn't work either.

 

This topic is rather pointless as there's no way to determine theoretical virtual rewards that will make every cacher content with the decisions that are made.  It makes for great debate about what should be considered if something were to take place, but in the end, there will still be people unhappy with the decision, regardless of the method that was used to hand out these types of rewards.

 

Popularity, as determined by FPs, doesn't necessarily denote quality, as others have said.  While the film can in the fish mouth was cute, I wouldn't consider it a high quality cache but I wouldn't consider it a bad cache either. It was just a cache.  It was different but the fact that it was in a parking lot of a mall prevented me from giving it a FP.  Place it next to a pond out in the woods with a nice hike to get there and it would receive a FP from me because it's a better fit there (my opinion, not a generalization), thematically.

  • Upvote 1
Link to comment
35 minutes ago, coachstahly said:

This topic is rather pointless as there's no way to determine theoretical virtual rewards that will make every cacher content with the decisions that are made.  It makes for great debate about what should be considered if something were to take place, but in the end, there will still be people unhappy with the decision, regardless of the method that was used to hand out these types of rewards.

I think the topic has an unintended point, which is that you cannot satisfy everybody. This thread can be pointed to in the future when people complain about the algorithms for determining future rewards (if there ever are any), to show that.

 

So bookmark this thread, to pull it out when the next round of complaints come in. B)

Link to comment
1 hour ago, coachstahly said:

Your argument for using current stats (which has evolved from current to a year long period of time) eliminates all caches that have been out for a few years, received FPs, but don't receive the same number of FPs they did originally because the locals have found them and they only get the occasional visitor now.  I fail to see how that's an equitable measurement for determining whether or not someone is in the running for a possible theoretical virtual reward.

 

It is quite normal to give awards for recent players but some awards goes to long career. Think about Academy Awards and Academy Honorary Award.

Link to comment
1 hour ago, coachstahly said:

How many people complained about the cachers who received a virtual reward who only had a few caches that had been placed?

 

I have no idea. I have not complained anything in public. I may have discussed about some selections and how they fit to the idea. A cacher with almost all caches archived or a quite new cacher who have only couple of caches and few hundred finds. It may relate to home coordinates or something else. I am glad that there are new virtuals - not complaining. Who is complaining?

Link to comment
16 hours ago, barefootjeff said:

 

I would hope that reaction time for DNF logs wouldn't become part of any algorithm. I just saw this DNF log on a cache on my watchlist:

 

 

Just how is a CO meant to respond to a DNF like that? Or two of them, if he'd logged a DNF for the first time he chickened out? Or several more if he brings some mates with him next time and they all chicken out?

 

No, of course responding to DNF logs shouldn't be part of anything. They're a natural, healthy product of good hides. Would you reexamine a test you designed if anyone failed to get a perfect score?

  • Upvote 1
Link to comment

Algorithmically, this is virtually impossible.


For example, take the issue of how they respond to NM logs, which I think I raised first.

 

Someone else said that you'd have to count that in the negative only, because if my caches get NM logs and you NEVER get one, then I've gotten an opportunity to garner more negative points (or whatever you're using to measure) than you.

 

That doesn't seem fair. Someone else suggested that a CO with no NM logs might score higher because they've created hides not prone to problems!

 

But, that doesn't work because the cause of NM logs is frequently, if not usually, completely outside of the CO's control: people.

 

So, scratch that, even if you normalize for the number of hides and terrain and remoteness and weather and seasons and number of finders and a plethora, veritable or not, of other factors.

 

Best throw out the whole criterion.

 

And, you can do this for most of the suggested criteria. Comparisons require much more of a common basis that most factors in cache-hiding can possibly provide because unless you're comparing LampPost Caches, or Tree Caches or something like that, there just aren't enough similar points to compare HIDERS.

 

Because of that, and I'd have to go back and refigger this, in no time we'd end up with a very small list of nothing.

 

Reminds me of.....and watch it all the way through.....    https://www.youtube.com/watch?v=G_Sy6oiJbEk

 

Link to comment
16 minutes ago, TeamRabbitRun said:

Algorithmically, this is virtually impossible.

 

Algorithm works but it will not give your predefined results. According to the OP, we are selecting 'worthy' cache setters for some reward. It happens that popularity is not a way to be 'worthy' if I have interpreted this conversation correctly. If some CO is not worthy for me I just skip those caches and go after other caches made by more worthy owners. For me, a favorite means that the cache was especially worthy. I have no idea how other players select worthy enough caches to go after. Do they use some algorithm on just pick a random cache?

Link to comment
7 hours ago, Team Microdot said:
7 hours ago, IceColdUK said:

All that said, I think they’re the best we’ve got to identify good / popular / interesting / unusual caches.  Absolute numbers shouldn’t be used, but as a percentage of PM finds (made since FPs were introduced) seems a reasonable measure.

 

I expect that the 'Wisdom of Crowds', over time, makes FP's a far more useful metric than some would have us believe.

 

I agree.  I spend a fair amount of time looking at various travel sites that have hotel/restaurant reviews.   One of things I look for is not only the average rating, but how many reviews there are.  If a hotel has an average 4 star rating (out of 5 stars), but only has 5 reviews, one more lower rating is going to drop the average significantly.   However, when a hotel or restaurant has hundreds or thousands of reviews (as is the case for many NYC restaurants/hotels), a 4 star rating is likely going to be pretty solid and not going to change much.

 

The same can be said for favorite points.  A cache with a small number of finds (and only PMs can add a FP) could have a 90% FP rating but that might not be as strong an indicator that it's a good cache than a cache with hundreds of finds and a 90% FP rating.   A larger sample size is going to provide more accurate results.

Link to comment

My algorithm
 

1) Has opted in to a reward system (Note: this only lasts 1 calendar year and you have to opt in again)

2) Premium member

3) Has never been suspended

4) Has logged at least 50 caches

5) Has logged at least 5 caches of the reward type

6) Has placed at least 1 cache

Why these 6? It shows a desire on the part of the cacher. This is also 100% within their control.
 

 

 

 

Link to comment
3 minutes ago, igator210 said:

My algorithm
 

1) Has opted in to a reward system (Note: this only lasts 1 calendar year and you have to opt in again)

2) Premium member

3) Has never been suspended

4) Has logged at least 50 caches

5) Has logged at least 5 caches of the reward type

6) Has placed at least 1 cache

Why these 6? It shows a desire on the part of the cacher. This is also 100% within their control.

 

Items 4 and 5 could be very hard to do in cache-poor areas.

Link to comment
6 hours ago, coachstahly said:

Popularity, as determined by FPs, doesn't necessarily denote quality, as others have said.  While the film can in the fish mouth was cute, I wouldn't consider it a high quality cache but I wouldn't consider it a bad cache either. It was just a cache.  It was different but the fact that it was in a parking lot of a mall prevented me from giving it a FP.  Place it next to a pond out in the woods with a nice hike to get there and it would receive a FP from me because it's a better fit there (my opinion, not a generalization), thematically.

 

Hah I found today a bat with a bison tube stuck in it's mouth. Had to laugh at the timeliness. No I did not give it a favorite point but it did give me a smile.

Link to comment
20 hours ago, Rathergohiking said:

Percentage of favorites. 

Variety of cache types. 

Variety of container types. 

Variety of cache sizes. 

Variety of D/T ratings. 

 

Anyone scoring high on those is highly likely to hide great caches.  Should focus on positive accomplishments not subjective behavior (response to DNFs, etc.) Just my opinion. 

Focus on positive things that are quantifiable and weight each item within each category. 

 

All of this discussion about response times to logs, etc is stupid in my opinion. There is no perfectly behaved cache owner that can be ascertained via an algorithm in my opinion. 

 

My stance is that hat if you use the basic categories I previously outlined I would expect pretty good results. 

Link to comment
18 hours ago, arisoft said:

 

Because you tends to give your favorites to your friends, not to good caches?

 

Can you give some examples of caches which have got many favorites recently but are not good caches.

 

I don't often give out favourite points at all, and I certainly try not to give favourite points to caches that already have a lot and I can't really see the reason for them.

 

Favourite points beget favourite points because people tend to gravitate towards caches with favourite points already, meaning that both they get more visits, and people often get swayed by the popular opinion and add more points.  This in itself is not a problem, until you start using that favourite points score as some kind of pseudo-objective measure.

 

No I'm not going to go naming caches that don't deserve favourite points, that's silly.  However, I think a good example many would agree with is Mingo.  As and of itself, it is a pretty lame cache with a so so maintenance record at times, and it gratuitously breaks numerous guidelines.  It gets tons of favourite points because it's the oldest active cache and the target of many a caching pilgrimage, and no other reason.

  • Upvote 1
Link to comment
5 hours ago, fizzymagic said:

I would start off with: 

 

Nobody who got all upset over not getting included last time. Especially nobody who whined about it in the forums.

 

That would exclude roughly 50 % of the people in this thread.  Which means my algorithm is by definition the best.

 

In that case I'd exclude all of those "poor winners" and/or jumped on some kind of perceived moral high-horse over it last time and basically insulted those who complained about it.  There goes the other 50%! :)

Edited by funkymunkyzone
  • Upvote 1
Link to comment
7 hours ago, funkymunkyzone said:

I don't often give out favourite points at all, and I certainly try not to give favourite points to caches that already have a lot and I can't really see the reason for them.

 

Favourite points beget favourite points because people tend to gravitate towards caches with favourite points already, meaning that both they get more visits, and people often get swayed by the popular opinion and add more points.  This in itself is not a problem, until you start using that favourite points score as some kind of pseudo-objective measure.

 

I have vague feeling that I may share your experience. Sometimes I give a favorite to a cache which have none or only few of them if I think that it is worth of more than it has been earned. But I do not agree that all other players just follow mindlessly what others do.

Link to comment
On 8/10/2018 at 3:31 PM, igator210 said:

My algorithm
 

1) Has opted in to a reward system (Note: this only lasts 1 calendar year and you have to opt in again)

2) Premium member

3) Has never been suspended

4) Has logged at least 50 caches

5) Has logged at least 5 caches of the reward type

6) Has placed at least 1 cache

Why these 6? It shows a desire on the part of the cacher. This is also 100% within their control.
 

 

Your "#6" - Remember, the OP said " If you were constructing an algorithm to select 'worthy' cache setters for some reward , ...", so if they only had one hide, it would have to be a seriously kick-a** one!

 

This is specifically about COs, not merely cachers.

Link to comment
2 minutes ago, TeamRabbitRun said:

Your "#6" - Remember, the OP said " If you were constructing an algorithm to select 'worthy' cache setters for some reward , ...", so if they only had one hide, it would have to be a seriously kick-a** one!

 

Good point. Why to punish a CO who have one extremely fabulous cache for 99 power trail micros he may also have.

Link to comment
1 hour ago, arisoft said:

 

Good point. Why to punish a CO who have one extremely fabulous cache for 99 power trail micros he may also have.

 

This attitude that anyone who isn't awarded is being punished is becoming rather tedious.

 

It would serve us all right if Groundspeak decided these award schemes weren't worth their trouble because no matter what they do someone will end up feeling 'punished'.

 

In the words of Elton John, it's a sad, sad situation.

 

I can't help thinking that some people just need to grow up and get over themselves.

Link to comment
1 hour ago, Team Microdot said:

It would serve us all right if Groundspeak decided these award schemes weren't worth their trouble because no matter what they do someone will end up feeling 'punished'.

 

In the words of Elton John, it's a sad, sad situation.

 

I can't help thinking that some people just need to grow up and get over themselves.

 

From what I can tell GS can't do anything without receiving criticism. Which is why they tend to do what they want to do and have their own evaluation for success/failure.

Link to comment
On 8/11/2018 at 7:53 PM, arisoft said:

 

I have vague feeling that I may share your experience. Sometimes I give a favorite to a cache which have none or only few of them if I think that it is worth of more than it has been earned. But I do not agree that all other players just follow mindlessly what others do.

 

Well wording it as following mindlessly might be a bit of an exaggeration.

 

But it's pretty clear that caches that stick their head slightly above others then get more and more visits, and perhaps because they are being targeted, that cachers tend to remember them more when they log them.

 

Take the most favourited cache in the world.  It's a fun and cunning hide for sure... but I've seen a thousand caches that were so much better in their cunningness, coolness, gadgety brilliance, boldness of hide, creativeness of camouflage, awesomeness of location, etc, etc...  Now it has been so much favourited, it's a must do cache on virtually anyone's list when they visit Berlin... and so the favourites cycle continues... :)

 

Link to comment
6 hours ago, funkymunkyzone said:

Take the most favourited cache in the world.  It's a fun and cunning hide for sure... but I've seen a thousand caches that were so much better in their cunningness, coolness, gadgety brilliance, boldness of hide, creativeness of camouflage, awesomeness of location, etc, etc.

 

I haven't seen more than a thousand I like better when compared to that cache.  Maybe close to 200, but certainly not much more than that.  I gave it a FP.  I still haven't found anything like it.  I DO remember my boys being more upset that the Lego store was closed when we visited. 

Link to comment
15 minutes ago, coachstahly said:

 

I haven't seen more than a thousand I like better when compared to that cache.  Maybe close to 200, but certainly not much more than that.  I gave it a FP.  I still haven't found anything like it.  I DO remember my boys being more upset that the Lego store was closed when we visited. 

 

But you cut off the most important part of what I was saying - that now it has lots of favourites, that makes it specifically a target cache for a visitor to Berlin, and therefore a magnet for more favourite points. Thus the cycle of favourite points begetting favourite points.

 

Link to comment
17 minutes ago, funkymunkyzone said:

But you cut off the most important part of what I was saying

 

Then why comment about how you've seen a thousand other caches that were better.  That detracted from your main point, which I tend to agree with, even though I don't think it applies consistently across the board.  It certainly doesn't for me.

Link to comment
1 hour ago, funkymunkyzone said:

 

But you cut off the most important part of what I was saying - that now it has lots of favourites, that makes it specifically a target cache for a visitor to Berlin, and therefore a magnet for more favourite points. Thus the cycle of favourite points begetting favourite points.

 

 

I can see what you're saying, but at the end of the day, even if you target high-FP caches, you can still only award FPs to 10% of them.  So long as percentages are used, rather than absolute numbers, I still think FPs are the best we've got.

 

BTW, I gave a FP to 'Lego - einer ist zuviel' too. ;-)

Edited by IceColdUK
Added a 'btw'
  • Upvote 1
Link to comment

Favourite points should not be seen as some indicator of objective or universal quality. It's a guide, an indicator that there's a good chance a person may enjoy it, and the more points the greater that chance - not that this is some rating scale with the highest favourite points being the best cache ever.  People need to stop interpreting the point system that way.

 

And agreed about this 'punished' mentality for not being rewarded. This is what led to the 'everyone gets an award' culture. You get punished (negatively) for doing something wrong. If you're not awarded something that relatively few people receive, whether for merit or by chance draw, it's not punishment, and the awarders shouldn't be accused of unfair punishment (which leads either to no on getting anything or everyone getting something, for fear of being taken to task for being unfair).  Even if there's a score system that doesn't seem favourable to you. It might feel unfair, but it's not punishment.  That doesn't mean the award system is perfect, unflawed in selecting awardees, but just this overuse of the term "punishment" - it's vastly misplaced.

  • Upvote 2
Link to comment
37 minutes ago, thebruce0 said:

Favourite points should not be seen as some indicator of objective or universal quality.

Exactly! I give a favorite point simply because it was ... one of my favorites. A lame, poor quality cache can still be one of my favorites (maybe not someone elses) for a number of reasons -- the scenic area, the accomplishment of finding it such as a high D/T cache, or whatever experience I went through to find it. Sometimes my favorite will be yours, other times not.

  • Upvote 1
Link to comment
4 hours ago, thebruce0 said:

Favourite points should not be seen as some indicator of objective or universal quality. It's a guide, an indicator that there's a good chance a person may enjoy it, and the more points the greater that chance

 

I agree with, with a caveat that if it's used to compare caches in a general area that it's a good indicator of the others may enjoy it as compare to other caches in the area.  Given that FPs can only be awarded by premium members that have found the cache, the relative number of geocaches in an area with available FPs to give can be significantly different depending on the area.  In places like many cities in Germany the frequency in which geocaches are found is much higher than in most other places.  The relative density of geocaches in the area also means that  cachers there can accrue FPs faster than those in cache sparse areas.   Comparing the number of FPs on a specific cache in Germany to a cache in Maine doesn't mean that the one in Germany would necessarily be enjoyed more than one in Maine, simply because geocaches are found far less frequently by geocachers with far less FPs to give than a cache in Germany.

Link to comment

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.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...