DGPT - Jonesboro Open

[Hoeschel]

OK.

I'll define it like this: If the two fields went head-to-head, which field would win? (Win is have the lowest score from either field.)

This takes into account the size and depth of the field, as well as the skills of the top players.

I simulated a 4-round tournament on a course with SSA=63, just based on player ratings. After a couple of thousand tournaments, the stronger field is the one that wins more often.

...

How do you model uncertainty in performance? If you only take the player rating, the best player should always win.

 

[lyleoross]

What I did was not a prediction. No two pools of players are ever going to go head to head. 1051 rated Paul from Vegas is not going to play 1054 rated Paul from Jonesboro. We will not be able to look at the results from Jonesboro and say whether the field was stronger than any other tournament or not.

Most of the discussion of stronger fields comes down to comparing the ratings of the players who are entered.

So, I made a model to approximate a more theoretically correct but enormously complicated way of using all those ratings to come up with a number that is bigger for one group.

Like I wrote, nerd sandwich. Best compliment I've ever given.

 

[Hoeschel]

Let's set politics aside. Statistical analysis, where you look at a subset of data, gathered carefully, is a pretty accurate method of determining final outcomes. No matter what Mark Twain said.

As someone working with statistics every day i have to disagree. You can arrive at vastly different conclusions using the same data but interpreting it differently.

 

[lyleoross]

As someone working with statistics every day i have to disagree. You can arrive at vastly different conclusions using the same data but interpreting it differently.

But that only proves the point. It isn't the numbers that changed, it's what the person wanted to see, how they looked at the numbers.

And the problem is that not all statistics work that way. Some are flat out, here's the numbers, here's the analysis, and there are standards as to the credibility of that result that are the norm. And folks are treating those as if they are the result of someone picking out the result they wanted.

When Lloyds uses stats to predict risk, they arent thinking, "What's the result I want." They are thinking, "What's gonna happen in the real world and can we accept that risk?"

The folks that are looking for a result are the problem, not the numbers. That brings us back to what I wrote earlier, lies, damned lies and the human beings who tell them.

BTW, you don't need statistics to tell a lie, just buck up to it and lie. Many politicians and sales persons do just fine with that model.

 

[lyleoross]

As an aside, I've seen actuaries fall prey to their humanity. They've developed a product that they want to succeed and they do poor analysis. Again, that isn't the numbers, it's their analysis and personal involvement that creates the problem.

 

[lyleoross]

How do you model uncertainty in performance? If you only take the player rating, the best player should always win.

Past performance can give you a feel. Is the person consistent or up and down. But there are too many factors imo.

 

[Flick Maniac]

Relax, I just wanted to use a saying my father is fond of. There are dozens, believe me.

Statistics dont lie. But you can pick and choose to suit your agenda, or interpret them in a handy way. But the underlying data is what it is, unless fabricated.

For example: a field of one Paul McBeth is an even match for a field of 100 A.J. Risleys. (Yes, I did the math.)

Now this is a Jomez production I want to see. Kinda like Neo vs the army of agent Smiths. <3

 

[Steve West]

How do you model uncertainty in performance? If you only take the player rating, the best player should always win.

For this, I just gave everyone a standard deviation (known from other work) and randomly selected their round ratings from a normal distribution with a mean of their round rating.

I think that's good enough for this model, since the only question was A>B and it is, after all, just a way to compare two sets of player ratings. And, it's just for fun.

If I were trying to predict the actual winner at the event, I would randomly select from the ratings-based scoring distribution for each hole for each player.

 

[Jolt]

Here we go, it´s live and i REALLY hope this is going to be a good live show. . .Discgolf need this to go well

 

[2naphish]

Tick Tock.....this could be epic.

 

[Jolt]

Lot´s of "blue" on the FPO cards. . have they made some of the holes to short for the FPO?

 

[lyleoross]

Question. I'm watching this on DGPT.tv, and am seeing no adds, is this the same for others? That is, have they blown the ads again? Hard to make money if you're not showing ads. Am I missing something?

 

[Jolt]

Ye, adds do not work. . again. . only the short "putt out" ones work

 

[Jolt]

So Henna plays in her first US tournament, she is on the lead card and on the live stream. . her sponsor is the main sponsor AND she have the owner of Prodiscus on the bag. . . WOW talk about pressure

 

[JC17393]

Question. I'm watching this on DGPT.tv, and am seeing no adds, is this the same for others? That is, have they blown the ads again? Hard to make money if you're not showing ads. Am I missing something?

I've seen some ads, but some breaks have been empty. I suspect that it is the ad-serving platform that is not functioning correctly rather than a lack of advertisers. And the split-screen ads during putt-outs are working because those are separately served. By that I mean they are the work of JVD in the production studio whereas the commercial breaks are not under his control (other than going in and out of them). Maybe JVD can check in between rounds and enlighten us.

 

[Halcón]

So Henna plays in her first US tournament, she is on the lead card and on the live stream. . her sponsor is the main sponsor AND she have the owner of Prodiscus on the bag. . . WOW talk about pressure

She's playing quite well. I wouldn't be surprised at all to see her win Worlds within a couple of years. And she's only what, 18 years old?

 

[lyleoross]

I've seen some ads, but some breaks have been empty. I suspect that it is the ad-serving platform that is not functioning correctly rather than a lack of advertisers. And the split-screen ads during putt-outs are working because those are separately served. By that I mean they are the work of JVD in the production studio whereas the commercial breaks are not under his control (other than going in and out of them). Maybe JVD can check in between rounds and enlighten us.

Yeah, I should have been clearer, I never thought there were no ads, just a problem with delivery. Steve has shown a great ability to get sponsors.

Note, I just saw my first ad. Someone needs to kill the into-exit beeps.

 

[Jolt]

I don´t think Henna or Eveliina have played worlds, i know Eveliina is hoping to play this year . . but hard for them to play in US

Bad last hole for Henna in the lead and on the last hole. . than that happens

 

[Halcón]

I really need to shut my mouth. I ruined Rebecca's win last week by saying she was playing great, and now I type up Henna's name and she triple bogey's hole 18. :wall:

 

[Steve West]

Lot´s of "blue" on the FPO cards. . have they made some of the holes to short for the FPO?

Too much blue would not indicate the hole is too short, it would indicate par is too high for FPO, relative to that length. Or, that conditions are extraordinarily favorable.

Too short would mean there is no scoring spread. None of the holes are close to being that short. In fact, the holes with the least scoring spread (3, 7, and 13) would have better scoring spread if they were made shorter so more low scores would better different the better players.

If conditions are ordinary, hole #9 should be an FPO par 4 in its current configuration. Holes #2 and #7 maybe should be FPO par 3. The other holes have good amounts of blue for this field.