Speed vs Spin

[sidewinder22]

Here's some Speed vs Spin data from the European Open. You can click on the player for more throws.

In general it seems Spin decreases as Speed increases. The lowest drive speeds tend to have the highest spin rates. Approach shots have higher spin rates than drives.

https://europeanopen.gameproofer.com/scoreboard/1


Thanks to NoseDownKing for the link!

 

[RFrance]

This might not be the best way to analyze the data but just quickly playing with it it does look like Spin decreases as Speed increases (averaging for each speed is probably better but it's late at night, so).

 

[BogeyNoMore]

Where were you when I started this thread 13 years ago?

https://www.dgcoursereview.com/forums/showthread.php?t=6378

 

[Sheep]

I want to know how they are measuring ripums on the discs.

 

[Sheep]

okay wow.

The average throw speed is pretty good.
But the variations in ripums is insane.
One guy throws 112km at 1020, then we see 115 with 2100 ripums

So, a quick off the cuff conclusion is, it's easier to generate speed/power than it is to generate spin.

HOLY CRAP.
134/2400
Insane!
Elias Gripler.
Thats 83mph.

 

[Sad2980]

Do you guys think there may be a lower threshold for spin necessary for a good throw? I.e. the disc must spin faster than 1000 rpms but anything higher than that is unnecessary? As a not very good Am, I've been wondering if I need more spin on the disc, but I know that my goal shouldn't be to increase spin above all else.

 

[seedlings]

Is it possible that the lower rpm throws are generally forehand?

 

[Brychanus]

Here's some Speed vs Spin data from the European Open. You can click on the player for more throws.

In general it seems Spin decreases as Speed increases. The lowest drive speeds tend to have the highest spin rates. Approach shots have higher spin rates than drives.

This might not be the best way to analyze the data but just quickly playing with it it does look like Spin decreases as Speed increases (averaging for each speed is probably better but it's late at night, so).

I'm really interested in these data & this relationship, glad they're finally coming out!!!

Though something interesting/funky going on unless I did something wrong. At the time I pulled the data just a bit ago, I got a different result than RFrance. I think the top 20 visible to me (attached from this analysis) are updating as they record new shots or something? They have the "on the tee line" which also appears to update (saw Kona up there, then Luke Bayne etc.)

When I plotted the data available at the moment I pulled it, I get the following distribution:

Notice that if you look over a wide interval in the X- and Y- dimensions (to avoid the "truncated graph" problem), at least at the moment I pulled the 20 available datapoints, this looks like a much less tight spread. The Pearson's correlation there is positive there, R = 0.15, or only 2% shared variance between Speed and Spin.

So I either (1) did something wrong, (2) there is a reliable negative or positive relationship, but we need all the data to be more confident or (3) as the data update, I suspect we might find that we're actually sampling from a distribution with a weak relationship between Speed and Spin, and some small N=20 data batches will show positive or negative relationships with varying slopes due to sampling error.

 

[twistedraven]

We need more data. Seems like a lot of outliers and a small sample size.

 

[Sheep]

the data being shown in the link was updating as people threw, and not holding any sort of rankings. just showing the last X players to throw.

 

[Skervoy]

Is it possible that the lower rpm throws are generally forehand?

Most likely yes, forehand is about half the spin normally. Which is also why they turn and fade more when throwing the same disc.

 

[sidewinder22]

This is wild. Not sure how accurate the data is, lot of guys throwing 80mph and one over 90mph. Some spin rates vary by 2000 rpm on same player within 1kmh at high speed. We don't know which throws are FH or BH or OH or roller.

 

[Brychanus]

We need more data. Seems like a lot of outliers and a small sample size.

This is wild. Not sure how accurate the data is, lot of guys throwing 80mph and one over 90mph. Some spin rates vary by 2000 rpm on same player within 1kmh at high speed. We don't know which throws are FH or BH or OH or roller.

Totally. I just PMed NoseDownKing if he knows more about full data access and am poking around for someone to contact about a full release/methodology. Interested enough in this to see if we can obtain something really meaningful.

 

[BogeyNoMore]

This is wild. Not sure how accurate the data is, lot of guys throwing 80mph and one over 90mph. Some spin rates vary by 2000 rpm on same player within 1kmh at high speed. We don't know which throws are FH or BH or OH or roller.

This is why different molds can behave so differently for different players. Even if two players have similar velocities, they may impart very different spin rates. This is a big part of what makes any given players throwing style somewhat unique.

As a result, a given disc will fly differently for different players, even at the same speed, line, trajectory, and angle.

Spin is what gives a disc stability, and holds it on line. A given disc requires a certain ratio of spin vs speed to achieve a relatively straight (i.e. stable) flight.

Understable discs require more spin to hold their line. Easy to throw them "too fast/not enough spin" ... this is why they are easier to turn.

Overstable discs require less spin to hold their line. Hard to throw them "too fast/not enough" spin. Easy to throw the with nit enough speed to achieve a truly stable flight, because they do t require much spin.

 

[txmxer]

Great concept for data collection. Need a LOT more data and details to start drawing meaningful conclusions.

 

[TractorFox]

I'm really interested in these data & this relationship, glad they're finally coming out!!!

Though something interesting/funky going on unless I did something wrong. At the time I pulled the data just a bit ago, I got a different result than RFrance. I think the top 20 visible to me (attached from this analysis) are updating as they record new shots or something? They have the "on the tee line" which also appears to update (saw Kona up there, then Luke Bayne etc.)

When I plotted the data available at the moment I pulled it, I get the following distribution:

Notice that if you look over a wide interval in the X- and Y- dimensions (to avoid the "truncated graph" problem), at least at the moment I pulled the 20 available datapoints, this looks like a much less tight spread. The Pearson's correlation there is positive there, R = 0.15, or only 2% shared variance between Speed and Spin.

So I either (1) did something wrong, (2) there is a reliable negative or positive relationship, but we need all the data to be more confident or (3) as the data update, I suspect we might find that we're actually sampling from a distribution with a weak relationship between Speed and Spin, and some small N=20 data batches will show positive or negative relationships with varying slopes due to sampling error.

Good observation (You didn't do anything wrong). I think it's important to keep in mind that the players they are showing *should* be very left skewed of the true distribution--they should be the strongest/fastest/tallest--whatever players. Which obviously introduces pretty intense selection biases.

For more data... If you click on the player's name you can get their previous throws as well as duration of flight and height of throw. At least then you can control for a few more of the factors and get an adjusted R^2 estimate.

Example: I just saw (Aapo Ojala Speed 86 km/h RPM 1980) and this person obviously isn't on the leaderboard.

Totally. I just PMed NoseDownKing if he knows more about full data access and am poking around for someone to contact about a full release/methodology. Interested enough in this to see if we can obtain something really meaningful.

This is pretty exciting, but you may want to temper your expectations for data sharing (I hope I'm wrong). They've gone to some lengths to protect the data from scraping (keys for players are randomly generated, the page loads dynamically, limiting to the top 20 throwers
). The host is a sensor company and probably has a vested interest in holding onto that data--maybe to do their own analysis?

 

[TractorFox]

I meant to add this. Here are Thomas Gilbert's individual throws so far
https://europeanopen.gameproofer.com/profile/b7d83143-812d-4761-9858-fe885ed80788

 

[Brychanus]

Good observation (You didn't do anything wrong). I think it's important to keep in mind that the players they are showing *should* be very left skewed of the true distribution--they should be the strongest/fastest/tallest--whatever players. Which obviously introduces pretty intense selection biases.

For more data... If you click on the player's name you can get their previous throws as well as duration of flight and height of throw. At least then you can control for a few more of the factors and get an adjusted R^2 estimate.

Example: I just saw (Aapo Ojala Speed 86 km/h RPM 1980) and this person obviously isn't on the leaderboard.



This is pretty exciting, but you may want to temper your expectations for data sharing (I hope I'm wrong). They've gone to some lengths to protect the data from scraping (keys for players are randomly generated, the page loads dynamically, limiting to the top 20 throwers
). The host is a sensor company and probably has a vested interest in holding onto that data--maybe to do their own analysis?

Nice, thanks - I wouldn't be shocked if we can't get full info either.

Agree re: selection bias. I'm now also curious about the within/between player variances in the spin/speed ratios even if the range is restricted to top players.

Hard to know what this will tell us without more info about the methods and throw types/discs etc., but I grabbed all the within-player data from the current 20 listed. I'll check later and scrape what I can if the player board updates.

 

[RowingBoats]

I'm having a hard time accepting these numbers as accurate lol. And if they are accurate, I have an even harder time understanding what we can glean from this.

 

[timothy42b]

If the variability of spin exists within backhand throws, and isn't thrown off (sorry) by the lower forehand spins, then might it cluster with types of swing? Wide rail vs conventional for example?

If the variability of spin exists within the same player throw to throw, then all bets are off, good data will be hard to find.