Help understanding player ratings

[SilentExodus04]

Wow, thanks for all the great information guys. I really learned a lot from it. I also learned that they don't do a very good job of explaining this on the official pages, so I appreciate you trying to do your best to answer this question that I'm sure comes up often in this forum.

It sounds like a bunch of smoke and mirrors if you ask me, LOL. I hope there is statistical math behind it, probably not worth trying to comprehend, but it feels so arbitrary to me and like anyone could 'change' the stats for any given round just because of a feeling they got, etc, or you know how hard the field played for the day.

It's just really interesting to me that it's not nearly as defined as I thought, and really much more flexible than I ever imagined.

It's cool to just give myself an idea of ok if I average par on most courses that at least means I'd be in the 800-900 range on average. I just love this game and want to learn more about it, so I'm not going to just "let it go" but I'm not going to laser focus on it either. I play to have fun, and I'm pretty competitive so just trying to figure out the next step of the game and I will probably start trying to check out a very low level league this year or perhaps a one off tourney to try to play in.

Really interesting to see all the different viewpoints and how much interest this thread brought though

Also thanks for the link to post 5 rounds and see an 'unofficial' ranking, that might be fun! I'll have to give it a shot and see what it spits out for me sometime just for fun.

 

[Cgkdisc]

No one needs ratings even those who want to play most PDGA events. But we know many competitive players do want PDGA ratings. You do get them as a result of playing PDGA sanctioned events. The thing is, we know how well you play on average in PDGA events will likely be worse than how well you play and how much easier the course plays in rec rounds. There's no easy conversion formula but we've found anecdotally that players who first play PDGA events seem to average 2-5 shots worse per round than what they think they average on that course in their rec rounds.

 

[JC17393]

It sounds like a bunch of smoke and mirrors if you ask me, LOL. I hope there is statistical math behind it, probably not worth trying to comprehend, but it feels so arbitrary to me and like anyone could 'change' the stats for any given round just because of a feeling they got, etc, or you know how hard the field played for the day.

Of course there's statistics behind it. The formulas are proprietary though, so it's not like they're going to publish all the nitty-gritty details for anyone to peruse (or steal).

 

[MikeK]

Of course there's statistics behind it. The formulas are proprietary though, so it's not like they're going to publish all the nitty-gritty details for anyone to peruse (or steal).

LoL. The formula isn't proprietary. The idea that ratings are complicated and require more than addition and division (4th grade math skills) is the greatest illusion ever sold to a group of disc golfers.

JC you're way too smart to not realize how easy it is to calculate ratings for any event with more than 10 rated players.

 

[Steve West]

Here's a performance-based way. It won't be very accurate, though. Especially for very high or vey low ratings.

Assuming you've been playing for a while, go out into a wide open field that is bigger than you can throw across. Pick something on the other side as a target. Try to throw as close to it as you can get.

Take a tape measure, measuring wheel, or laser range finder (not an app or GPS device) and measure the distance from where the disc landed back to the spot you threw from.

Repeat 30 times and average the results. Include every throw, even the ones you didn't throw well and the ones that had great skips or rolls.

If you are male and averaged 300 feet, figure you are rated about 800. Add (or subtract) 2.6 ratings points for every foot over (or under) 300.

If you are female and averaged 240 feet, figure you are rated about 800. Add (or subtract) 1.4 ratings points for every foot over (or under) 240.

Disclaimer: Ratings are NOT based on how far you can throw, though more highly rated players do tend to throw farther. That's what this is based on. Don't take it seriously.

For me, this formula is off by about 200 ratings points. I'm apparently the shortest thrower in the world for my rating.

 

[DavidSauls]

Steve, that formula would miss me by about 200 points, too. I'll have to challenge you to a short-throwing rated round to see which of us deserves the title.

 

[DavidSauls]

Wow, thanks for all the great information guys. I really learned a lot from it. I also learned that they don't do a very good job of explaining this on the official pages, so I appreciate you trying to do your best to answer this question that I'm sure comes up often in this forum.

It sounds like a bunch of smoke and mirrors if you ask me, LOL. I hope there is statistical math behind it, probably not worth trying to comprehend, but it feels so arbitrary to me and like anyone could 'change' the stats for any given round just because of a feeling they got, etc, or you know how hard the field played for the day.

.

It would be nice if the PDGA included a simple summary, a very general description introduction in simple language (without the word "propagator" to be found). Then went into the details, as far as they're going.

The basics are simple. The finer details may not be, but the basics might go something like this (subject to my hasty writing, before coffee has kicked in)"


---Your round rating is based on how you do, compared to everyone else in that round.
---The players in a tournament bring their pre-existing ratings.
---Based on those ratings, and their scores in a round, a benchmark is set for that round. (If the group has an average rating of 950 going in, and their average score is 50, then 50 will be a 950 rated round).
---Other scores produce ratings by how much they vary from that benchmark.
---A rating for an individual round is imprecise. But when a player averages his or her round ratings from dozens of rounds, they get a player rating in which the variances of individual rounds even out.
---The purpose of the rating system is to produce player ratings, and thus allow Ams to play in divisions with similarly-skilled players. It is not to produce a specific rating for a single round, that has any great meaning.


But better. And, yes, when you get the details (like which players' scores contribute to that benchmark), it gets more complicated. But something like that would be a start, for beginners trying to make sense of the ratings system.

 

[MikeK]

David you might want to add "round ratings have absolutely nothing to do with the course par or course rating."

 

[DavidSauls]

David you might want to add "round ratings have absolutely nothing to do with the course par or course rating."

Good point. Just avoiding the word "par" probably isn't enough.

The trick is saying enough to give newcomers a general idea, without bogging down into discussions on weather, hot rounds, the effect (or non-effect) of high-rated players, why ratings on the same course vary from round to round, etc. A Q&A might follow---if there's not already one, I don't know, I haven't had a need to look.

The basic concept is simple. My wish is for a simple introductory explanation, that we can point new players to.

Someone smarter, and less verbose, than me should do it, though.

 

[Steve West]

Steve, that formula would miss me by about 200 points, too. I'll have to challenge you to a short-throwing rated round to see which of us deserves the title.

At Am Worlds?

 

[DavidSauls]

Doesn't look like I'll make another Am Worlds, at least not in the foreseeable future. The twin hurdles of playing enough events to get in, and being able to commit a week to it. Too much Other Life in my life, these days.

 

[JC17393]

LoL. The formula isn't proprietary. The idea that ratings are complicated and require more than addition and division (4th grade math skills) is the greatest illusion ever sold to a group of disc golfers.

JC you're way too smart to not realize how easy it is to calculate ratings for any event with more than 10 rated players.

Sure. But if you're the PDGA and you put ratings out as one of the great benefits of membership, are you going to publish a step by step tutorial on how to calculate them? If people want to reverse engineer the numbers and figure out the calculations for themselves, great (I agree they're not complex).

I was only responding to the suggestion that the PDGA doesn't explain ratings and how they're derived well enough. They have nothing to gain by doing so.

 

[TimSyl]

I assume this is as good a place as any to put this. Had a tournament this past weekend at a complex, Meyer Broadway, with 2 18 hole courses. AM2 was the only group to play longs on the North course in the morning. A guy shot 53, and it was rated at 924. Pros and AM 1 played the north in the afternoon under basically the same conditions. A 53 in the afternoon was 974. Huh?????????

 

[Cgkdisc]

I assume this is as good a place as any to put this. Had a tournament this past weekend at a complex, Meyer Broadway, with 2 18 hole courses. AM2 was the only group to play longs on the North course in the morning. A guy shot 53, and it was rated at 924. Pros and AM 1 played the north in the afternoon under basically the same conditions. A 53 in the afternoon was 974. Huh?????????

Whenever you see that big of a difference, there's a good chance the course layouts didn't get assigned properly or perhaps TD hasn't gotten around to it. This is common in events where more than one set of tees or more than one course are used.

 

[VictorB]

I assume this is as good a place as any to put this. Had a tournament this past weekend at a complex, Meyer Broadway, with 2 18 hole courses. AM2 was the only group to play longs on the North course in the morning. A guy shot 53, and it was rated at 924. Pros and AM 1 played the north in the afternoon under basically the same conditions. A 53 in the afternoon was 974. Huh?????????

Playing at a different time could mean:
1. many more propagators on the course in the afternoon
2. conditions weren't exactly the same, even if they were close
3. smaller scoring spread by propagators
4. pool didn't get assigned the right tees in the initial score upload

It's basically impossible to really compare a round played at a different time with different pools, since that's not what the system was really intended for

 

[DavidSauls]

I'm betting on the TD's layout assignments, unless there was a small number of Am2s, so that a few atypical rounds had a major effect.

 

[JC17393]

I'm betting on the TD's layout assignments, unless there was a small number of Am2s, so that a few atypical rounds had a major effect.

I'm with Mr Sauls, particularly the smaller number of players/propagators in AM2 vs the pool with Pro and AM1 players. The smaller pool of propagators there is, the bigger influence an individual round can have on the numbers.

 

[TimSyl]

Whenever you see that big of a difference, there's a good chance the course layouts didn't get assigned properly or perhaps TD hasn't gotten around to it. This is common in events where more than one set of tees or more than one course are used.

I was the TD and assigned it properly.

 

[TimSyl]

It was a small AM2 sample size. This is/just an example of my questioning the system. I find it strange you have to have high rated players to get a high rating.

 

[JC17393]

It was a small AM2 sample size. This is/just an example of my questioning the system. I find it strange you have to have high rated players to get a high rating.

Not true. 100% myth.

While it is more likely that lower rated propagators have ratings that lag behind their skill level (as they rapidly improve), that doesn't mean that higher ratings automatically equal more reliable or "accurate" ratings. Higher rated players tend to be more stable (their rating isn't lagging behind their rate of improvement much if at all), but there are plenty of lower rated players whose rating is just as stable. A pool of stable 800 rated players should generate near identical (within a few points) ratings to a pool of stable 1000 rated players on the same course in the same conditions. The key word is stable.

The ratings from the AM2 field are more likely skewed disproportionately by one or two specific players who played well above their current rating. Individuals who wouldn't have the same impact shooting exactly the same way in a larger field. If you had 10 propagators, that means each of those players accounts for 10% of the calculation. If you have 50 propagators, everyone accounts for 2% of the calculation. That's a big difference. More propagators means the oddball or two that significantly outplay their rating have a much lesser (negligible, even) impact on a particular round's ratings.