DGCR v UDisc Rating

[bigreddog]

Anyone who has used both DGCR and UDisc knows that UDisc ratings are inflated by comparison. I recently did some data analysis to see just how inflated and I think other people would be interested in seeing it.

I took the 137 courses that I have played that are not "RIP"-ed and recorded the DGCR Rating as well as the UDisc rating for all of them. I then fit a line to the data (shown below)




If the ratings were not inflated, we would expect the trendline to have a slope of 1, and an intercept of 0. What we actually see is a slope of about 0.5 and an intercept of about 2.4. This means 2.4 is about as low as UDisc ratings go, so essentially they only use about half of the scale (2.4 - 5.0). DGCR on the other hand uses pretty much the entire scale, there are courses with 5+ reviews rated 0.1, with 22 courses with 5+ reviews rated 0.75 or below.

I really wish DGCR was the standard tool that everyone used, rather than UDisc. I get that people like UDisc for tracking their round, but the course information provided is just so much more meaningful on DGCR.

 

[bigreddog]

If anyone would like to assist with adding more data, here is a link to the google sheet where I did the data analysis. Just scroll to the end and you can input course name, DGCR rating, UDisc rating.

usp=sharinghttps://docs.google.com/spreadsheets/d/1v0e7ie9dc38zMDO_WP_ww-wT9PhzhIGNTkmW9DSHTFI/edit?usp=sharing

 

[Ahildy13]

Anyone who has used both DGCR and UDisc knows that UDisc ratings are inflated by comparison. I recently did some data analysis to see just how inflated and I think other people would be interested in seeing it.

I took the 137 courses that I have played that are not "RIP"-ed and recorded the DGCR Rating as well as the UDisc rating for all of them. I then fit a line to the data (shown below)




If the ratings were not inflated, we would expect the trendline to have a slope of 1, and an intercept of 0. What we actually see is a slope of about 0.5 and an intercept of about 2.4. This means 2.4 is about as low as UDisc ratings go, so essentially they only use about half of the scale (2.4 - 5.0). DGCR on the other hand uses pretty much the entire scale, there are courses with 5+ reviews rated 0.1, with 22 courses with 5+ reviews rated 0.75 or below.

I really wish DGCR was the standard tool that everyone used, rather than UDisc. I get that people like UDisc for tracking their round, but the course information provided is just so much more meaningful on DGCR.

DGCR is a Forum, I could honestly care less about the ratings portion of this website. I would assume I am not in the minority.
Udisc asks me to rate a course after I play, however I usually don't.

If I want to see if a course is any good, I just go play it.

 

[BruisedOoze]

I think part of the problem, aside from the everyone has input aspect, is that Udisc captures a lot of people that only play a small handful of courses. They're likely still having fun and that is likely the only factor driving their evaluation.

I think knowing the transformation values you provided might help us eyeball the quality of a course based on Udisc scores alone.

Even within DGCR, eye-balling the data, it seems 3.5 seems to be about the central mode of the data and close to the average. This still bothers me a bit where, under ideal circumstances, 2.5 should be average. Then again, is the goal to determine what's average, or what the general enjoyment factor is? Not sure, and I currently don't have strong feelings on it yet.

 

[Ahildy13]

Even within DGCR, eye-balling the data, it seems 3.5 seems to be about the central mode of the data and close to the average. This still bothers me a bit where, under ideal circumstances, 2.5 should be average. Then again, is the goal to determine what's average, or what the general enjoyment factor is? Not sure, and I currently don't have strong feelings on it yet.

There is no reason the average rating should be 2.5, there is also no reason the lowest rated course couldn't be a 2.5.
Since these are ratings, a higher average simply means that we have good quality courses. If every course was "perfect" the average rating would be 5 and we would live in a DG utopia.

 

[nothinbuttree]

I think part of the problem, aside from the everyone has input aspect, is that Udisc captures a lot of people that only play a small handful of courses. They're likely still having fun and that is likely the only factor driving their evaluation.

I think knowing the transformation values you provided might help us eyeball the quality of a course based on Udisc scores alone.

Even within DGCR, eye-balling the data, it seems 3.5 seems to be about the central mode of the data and close to the average. This still bothers me a bit where, under ideal circumstances, 2.5 should be average. Then again, is the goal to determine what's average, or what the general enjoyment factor is? Not sure, and I currently don't have strong feelings on it yet.

There is a built in bias though for DGCR users. We read reviews here, and if a course seems like a dog, we will skip it, baggers notwithstanding. I will never review a course I never play. So other than bagging my local area, I'm very unlikely to play low rated courses out of town. I don't think udisc users are quite that discriminating in general.

 

[Monocacy]

Even within DGCR, eye-balling the data, it seems 3.5 seems to be about the central mode of the data and close to the average. This still bothers me a bit where, under ideal circumstances, 2.5 should be average.

I would expect the average review score to be higher than the average course rating, because better courses attract more reviews. Similarly, since the mode is the number that occurs most often in a data set I would expect the mode to be higher than the average course rating.

But I am not looking at the numbers so my speculation is unverified.

 

[Pevio]

There is a built in bias though for DGCR users. We read reviews here, and if a course seems like a dog, we will skip it, baggers notwithstanding. I will never review a course I never play. So other than bagging my local area, I'm very unlikely to play low rated courses out of town. I don't think udisc users are quite that discriminating in general.

The average review should certainly be above 2.5 because of selection bias. However, the average course should be around 2.5 (maybe not exactly, but somewhere around it). in other words, a bad course will have one or two low reviews, while a good course will have many high reviews.

 

[wellsbranch250]

DGCR is a Forum

huh?
you do realize DGCR stands for "disc golf course REVIEW"
DGCR has a forum. it is not just a forum.

 

[Cgkdisc]

It would be interesting to check the data if P2P and private courses score higher on UDisc than free courses where both have the same DGCR rating. It goes against human nature to continue to play a course you give a low UDisc rating unless you're a masochist with no other courses to play. Most are going to give a good rating to courses they play regularly and likely even higher if they are paying and know the owner to simply indicate their good taste regardless of the course's intrinsic quality (if there really is such a thing.)

 

[bigreddog]

Even within DGCR, eye-balling the data, it seems 3.5 seems to be about the central mode of the data and close to the average. This still bothers me a bit where, under ideal circumstances, 2.5 should be average. Then again, is the goal to determine what's average, or what the general enjoyment factor is? Not sure, and I currently don't have strong feelings on it yet.

I think the average on DGCR is lower than 3.5, probably closer to 3. There are hardly any 0's given though so, DGCR is closer to a 0.5 - 5.0 scale. With that you would expect 2.75 to be average, but you also want to keep in mind that low quality courses are more likely to be "RIP"-ed which would bump the average up a bit.

 

[Ahildy13]

huh?
you do realize DGCR stands for "disc golf course REVIEW"
DGCR has a forum. it is not just a forum.

I 100% understand that, but for me, and I am guessing a vast majority of users, it is just a Forum.

 

[BillFleming]

huh?
you do realize DGCR stands for "disc golf course REVIEW"
DGCR has a forum. it is not just a forum.

Yes, it isn't JUST a forum, but I wonder how many people on here rate/review courses compared to just posting/reading the forums? I'm here for the forums....that's what is important to me.

I don't review/rate courses on any app....ratings are subjective to how I feel at the moment and how the course played at that moment. It's like disc flight numbers...they aren't the same across brands....course ratings aren't the same across players. My 3 might be another player's 2 or 4 (or something else). Until numbers are defined as to what they mean, where there is no gray/grey area,....there's going to be discrepancies in ratings just in one app, much less all.

I mainly play local courses that I know. If I'm playing elsewhere and there are multiple courses to choose from, I might look up ratings/reviews to see which course I play....otherwise, I play and enjoy what is available. I've played courses that were "terrible", based on other player's opinions or reviews, and found them to be enjoyable. Likewise, I've played "enjoyable" courses that I thought were crap and would never play again.

 

[DatRedDude]

to rate a course on DGCR its take a little bit of effort and thought, on udisc you click a star rating with very little effort. i trust reviews her more than there but i believe the population as a whole uses udisc more than DGCR.

 

[Ahildy13]

I think the average on DGCR is lower than 3.5, probably closer to 3. There are hardly any 0's given though so, DGCR is closer to a 0.5 - 5.0 scale. With that you would expect 2.75 to be average, but you also want to keep in mind that low quality courses are more likely to be "RIP"-ed which would bump the average up a bit.

Again, there is no reason that you should have any expectation of the rating for the "average" disc golf course. It is mathematically meaningless, as there isn't only one course that is the "perfect" 5.0 and one course that is the "worst" 0.0.

 

[Countchunkula]

I 100% understand that, but for me, and I am guessing a vast majority of users, it is just a Forum.

I was always under the impression that more people used the dgcr course directory than the forum. The course directory with reviews is the stated purpose of the site. The forum is just an add on. We're both just speculating though, so neither impression is worth much.

 

[bigreddog]

I 100% understand that, but for me, and I am guessing a vast majority of users, it is just a Forum.

Maybe a majority of forum users, which is what you would see if you only use the forum, but I don't know about a majority of total users of the site. I only recently started looking at the forums but I've written reviews on and off since I started playing in 2012, so for me I only cared about the review portion.

 

[Ahildy13]

I was always under the impression that more people used the dgcr course directory than the forum. The course directory with reviews is the stated purpose of the site. The forum is just an add on. We're both just speculating though, so neither impression is worth much.

The most recent review was posted 10/23, before that 10/16.
There are currently 333 people browsing the forums.
I'd say the data is in my favor.

 

[nothinbuttree]

The average review should certainly be above 2.5 because of selection bias. However, the average course should be around 2.5 (maybe not exactly, but somewhere around it). in other words, a bad course will have one or two low reviews, while a good course will have many high reviews.

You mentioned courses with 5+ reviews so I thought that was part of your criteria. My bad if I read that wrong. You are correct in what you say above.

 

[grube.fresh]

Your "data" is incorrect.