Fangraphs for disc golf?

[Moose33]

Hey guys,

I'm a bit of a stat head, and mostly in baseball which is the only sport I am more heavily invested in that DG. One of the things that I really enjoy is the scouting, statistics and player development side.

One of the best sites for this is the popular site fangraphs.com which has the most in-depth stats and advanced analytics, but it also has a "fans scouting report section". In this section all fans grade players individual skills on a 20-80 scouting scale, which is the most common way to communicate individual skill values.

example of the site in link; http://www.fangraphs.com/statss.aspx?playerid=10155&position=OF

80 would be the best of the best like Simon Lizotte would get an 80 in drive distance for example, and 20 would be poor. A 40 is pretty baseline average. Usually you rank in increments of 5.

I would like to try something like this for disc golf, and since the people that post here are usually more heavily invested in watching players and have played with many of them I think with the right data we can get a pretty good picture of what players have the best skills or "tools" in our young sport.

One thing that I am trying to ascertain is; what do you need to win in DG?

is it enough to have one 80 skill as long as the rest are above 40? Do you need to have most of your skills in the above average range(50+).

How good do your individual skills have to be to be a top flight disc golfer.

So basically I have two things to ask of you;

One- what skills, that are purely physical and therefore less subjective should we try to grade?

Two- will you help assign grades to every pro player that you feel you can accurately scout?

These are the categories I have in mind so far;

Backhand power;
Forehand power;
Backhand accuracy;
Forehand accuracy;
Specialty shots;
Putting inside the circle;
putting outside of the circle;
upshots and approaches;


So say the player is Paul McBeth, whom most people who watch have seen play would be something like this for example;

Paul McBeth
Backhand power;70
Forehand power;65
Backhand accuracy;75
Forehand accuracy;70
Specialty shots;60
Putting inside the circle;80
putting outside of the circle;75
upshots and approaches;75


don't worry if you agree with those or not it's purely for an example on how it would look.

Let me know what you think, and if enough people participate I'll post my spreadsheets.

I'd really like to get the top 20 or so MPO and 10 FPO players rated with enough data to support findings and remove outliers(top 5 and bottom 5% probably, in line with how they do on fnagraphs)

Thanks!

 

[discgolfer25]

The new stats from the DGPT and DGWT should help with this. Check them out. I know they have putting percentages, fairway hits percentage, parked percentage, etc.

 

[Moose33]

Those stats are amazing, but I'm looking to do something like the scouted by the fans section, which is something that doesnt currently exist.

If the users look at that data, it will make it much better, but it doesn't tell 100% of the story.

 

[ohtobediscing]

You need to pad the stat categories, with things like

Head in the game
Rollaway distance
Basket reject %
# of gear/clothing failures
Average blood pressure

Now we're getting somewhere.

 

[hellbound]

How can you define things like Backhand power? Raw Open field power vs. Real Game play. Scouting using baseball as a reference uses radar guns, fixed field measurements. Not all pro's ever really throw for full on 90% of the drives. Heck I would go as far to say not vary many of the top guys even tried the distance comp at worlds.

I personally like the idea and Hope that Udisc stats become the norm vs SpinTV's. Intersting Idea but wondering how metrics would be set at for a base number.

 

[StevenSeagal]

Somewhere in there I think it could be useful to quantify how skilled an individual is in a scramble situation. Udisc describes scramble as missing circle 2 in regulation but still making par or better.

 

[hellbound]

Somewhere in there I think it could be useful to quantify how skilled an individual is in a scramble situation. Udisc describes scramble as missing circle 2 in regulation but still making par or better.

I agree with this.

 

[dwindacatcher]

First off, 50 is league average.

Second, You are giving mcbeth WAY to high on his skills. A 75 in "outside the circle putts" pretty much means if he is withen 50 foot he will make 29 of 30. That just isn't realistic. A 70 in baseball is still an elite talent. A guy with a 70 hit tool is going to bat over 300 ever year.

Hell, I'd probably give lizotte's power a 70. A guy with 80 power should be able to park 550 foot uphill turnover shots.

But arguing points aside, there isn't enough scouting in disc golf to reliably make these grades. I mean, what is the average touring players power? What does the average player make in putt % inside 40 feet? You could probably either find it or do the work to find it for the top 10, but what about whomever is ranked 107? His stats would have to be accounted for the average.

 

[Moose33]

I put numbers to show how to do it, not how to argue points. I would have to look at anyone's game more in depth to give my own metrics.

I know we don't have good scouting in disc golf, that is why this excercise is nessicary. We need to make an attempt to start getting better in that part of our game.

If you have opinions on how a person should be rated please post them! If love to start gathering data.

Most of the posters on fangraphs are in no way qualified to do so but in most cases with a big enough sample size of data it works out pretty close most of the time. We just need a big data pool to make it work.

In response to some of the other answers in relation to stats, I'm very interested in those as well but I feel like there are people progressing on those avenues.

No one to my knowledge is evaluating players based upon their individual skills. This is something that I believe correlates to the stats and gives a better overall picture of who the player is.

 

[BrotherDave]

I have no idea about any of this but Schwebby might be an 80 in overhead shots so there's my contribution.

 

[dwindacatcher]

Well, I have a thought for what to track. I don't think a forehand or backhand should be broken down to power and accuracy for a basic scouting report that the 20-89 scale is. I would suggest these 4. Forehand, backhand, putting, and approach/upshots.

Baseball uses 5, but I'm torn on a 5th. Best I've thought was rollers, but that could be part of forehand/backhand. But than again a roller is very different.

 

[onemilemore]

I put numbers to show how to do it, not how to argue points. I would have to look at anyone's game more in depth to give my own metrics.

I know we don't have good scouting in disc golf, that is why this excercise is nessicary. We need to make an attempt to start getting better in that part of our game.

If you have opinions on how a person should be rated please post them! If love to start gathering data.

Most of the posters on fangraphs are in no way qualified to do so but in most cases with a big enough sample size of data it works out pretty close most of the time. We just need a big data pool to make it work.

In response to some of the other answers in relation to stats, I'm very interested in those as well but I feel like there are people progressing on those avenues.

No one to my knowledge is evaluating players based upon their individual skills. This is something that I believe correlates to the stats and gives a better overall picture of who the player is.

It might be fun to try and poll some of the current players and have them give numbers on their colleagues.

 

[bobmcnelly]

I don't mind starting this off....but I've only watched one top player in the world play in person and watched online. Eagle Mcmahon.

Eagle Mcmahon
Backhand power;75, I'm sure he'll be 80 soon if he's not already.
Forehand power;60
Backhand accuracy;70
Forehand accuracy;65
Specialty shots;60
Putting inside the circle;65
putting outside of the circle;65
upshots and approaches;70

 

[Moose33]

It might be fun to try and poll some of the current players and have them give numbers on their colleagues.

That would actually be ideal since pro players have the best idea of a baseline and have seen more top players preform than anyone else.

If anyone would like to contribute data but not post publicly feel free to PM me.

Thanks!

 

[Cgkdisc]

I think there's a more effective way to determine skill sets of players but it's not by the Fangraph approach. With PDGA round ratings, you have a summary of all skills a player executes to produce a score on a specific type of hole or course, under specific conditions. Unfortunately, the hole and course characteristics plus conditions in each round are not codified and captured yet. But if it were, then regression analysis would be able to parse out how well a player scores in the wind, with elevation, avg hole length, under tournament pressure, in close finishes, playing second round in a day, in the woods, in the open, with lots of hazards, etc. Once that's codified for each player then just apply that to the characteristics of weather and the next courses being played in a tournament.

 

[Moose33]

I agree with that to some extent Chuck, but we have in my opinion too many factors that are out of a players control and that actually tell you nothing about the players individual skills. For example Simon is an amazing power thrower he would be a true 80 on the backhand power scale, but based on his stats you may not see that since he often misses fairways and throws can be more adversely affected by outside conditions.

My hope is that if you see that for example Ricky has a 70 rating in outside the circle putting but he was making only 25% of his outside the circle putts you can, without having seen the round decipher that he is better at that skill than he showed during the event.

Maybe he was dealing with an injury, or the wind was overwhelming.

By combining an approach with more accurately maintained stats, along with a large sample size scouting approach we can get the best evaluation of a players skill set.

I know you are the ratings guru, but to the average fan or even the more intense ones, it's too nebulous to get a feeling for individual skill sets for players. I think it's a good overall idea of TTL, but it doesn't say for instance that Paige Pierce has better overall backhand power than Valarie Doss.

My issue with a condition bases analysis is that our sample size over a given time period is too small to get a good picture of what the players actual abilities are.

Also, if we were ever able to make a disc golf video game featuring the actual players, it would be easier for developers to make accurate representations of our players in that context as well.

I've listened to many of the podcasts you have been on and I think you have a better take on what is going on that almost anyone else, but there are things that need to be debated to get the overall best result for our sport. Not that I am claiming to have all the answers. Just one small piece of the puzzle that makes sense to me.

 

[Cgkdisc]

Bottom line is what stats approach does the best job of forecasting real world performance. The process I'm proposing incorporates all of the elements based on aggregate probabilities without parsing out detailed skill sets. The stats being gathered now by both tours are not yet useful because they are not tied to hole and weather characteristics. For example, skills like driving power/distance, green hits and approach stats are meaningless outside of environmental connections.

 

[Moose33]

That is something we 100% agree on. And the data sets for both are far too small to tell us much of anything we didn't' already know yet.