Dr. Dave’s VNEA Vegas Experience

mikepage said:
It is not that it is not a valid match in some way. But if Fedor beat him 90-60, there wouldn't be a post asking if the games will go into FargoRate. And for the match to be statistically valid, it would need to go in regardless of the score. It cannot go in preferentially when the score is interesting (like here) or when the score is support that somebody you think is underrated is better than his rating.
Click to expand...

It has nothing to do with who won and how much they won by. You have data available to use and your choosing not to use it for some reason. Whether or not these games were being submitted to Fargo had no effect on the outcome!
 
Last edited:
  • Like
Reactions: SEB
nicksaint26 said:
It has nothing to do with who won and how much they won by. You have data available to use and your choosing not to use it for some reason. Whether or not these games were being submitted to Fargo or not had no effect on the outcome!
Click to expand...
Again, the issue is not whether the inclusion affects the outcome, it's whether the outcome affects the inclusion.
 
u12armresl said:
Why are the people who talk most about huge buckets, hitting the rail first (still making the ball), and the associated, not putting up large packs every time they get to the table.
Click to expand...
Just because you can tell a pocket is 5.5 inches wide doesn`t mean you can play.
 
dr_dave said:
FYI, I just posted a video shows some highlights from my recent visit to Vegas for the annual VNEA League Championship Tournaments.



As always, I look forward to your feedback, comments, questions, complaints, and requests.

Enjoy!
Click to expand...

Thanks for the update Dave, very well written. It's a shame it got derailed so quickly.
 
book collector said:
Thanks for the update Dave, very well written. It's a shame it got derailed so quickly.
Click to expand...

Actually, I think the pocket size and FargoRate discussions have been very interesting, much more interesting than my “Vegas Experience.” Although, I am a little surprised my new jump shot video thread didn’t generate more discussion, especially concerning the unintentional scoop rule issue.
 
nicksaint26 said:
The outcome doesn’t matter the data does.
Click to expand...
I'll take a stab at it. Here's a hypothetical situation where all of the data is completely valid but the result is crooked.

I have a great table at home and I like to bet high. Five of my friends are pretty close to me and we regularly play races to 5 for $200 to $500. All of the sets are on the up-and-up with everyone trying to win. I send the matches in to FargoRate since we all have ratings and I provide a link to a video of each match so FR can see that all the data is perfectly valid. So far, so good.

The catch is that I only send in the sets where I either lose or win hill-hill. I omit the sets where I dominate. My buddies can't understand why my rating is so far below theirs.

The problem is that the decision to add the data is made after the data is available. That is a huge problem.
 
Bob Jewett said:
I'll take a stab at it. Here's a hypothetical situation where all of the data is completely valid but the result is crooked.

I have a great table at home and I like to bet high. Five of my friends are pretty close to me and we regularly play races to 5 for $200 to $500. All of the sets are on the up-and-up with everyone trying to win. I send the matches in to FargoRate since we all have ratings and I provide a link to a video of each match so FR can see that all the data is perfectly valid. So far, so good.

The catch is that I only send in the sets where I either lose or win hill-hill. I omit the sets where I dominate. My buddies can't understand why my rating is so far below theirs.

The problem is that the decision to add the data is made after the data is available. That is a huge problem.
Click to expand...

'reporting bias'. One of the things Elizabeth Holmes is in prison for.
 
Bob Jewett said:
I'll take a stab at it. Here's a hypothetical situation where all of the data is completely valid but the result is crooked.

I have a great table at home and I like to bet high. Five of my friends are pretty close to me and we regularly play races to 5 for $200 to $500. All of the sets are on the up-and-up with everyone trying to win. I send the matches in to FargoRate since we all have ratings and I provide a link to a video of each match so FR can see that all the data is perfectly valid. So far, so good.

The catch is that I only send in the sets where I either lose or win hill-hill. I omit the sets where I dominate. My buddies can't understand why my rating is so far below theirs.

The problem is that the decision to add the data is made after the data is available. That is a huge problem.
Click to expand...
This is a problem with any rating system. Sandbagging can be done somehow. Most people won’t do it. Going from memory I think the FargoRate site talks about it and there is generally a cost to doing so i.e. paying to enter and dumping. The golf handicapping system is subject to it because scores are generally self reported. A couple things slow it down. Handicap committees and tournament scores being noted differently. Again, different because of the stroke play v. Match play issue. Shoot a couple 76s in a tournament as a 14 and your handicap may go down. Golf they can do some math and figure the likelihood of a given handicap shooting a given score. There was a recent article in Golf Digest about it.

In pool I can see the issue for players who are really better than a given rate keeping their Fargo below the cut off for a given tournament. Eventually they will be figured out. For instance the tournaments with a 720 cut off. Can you tell the difference visually between a 731 and 714? Probably not. If a player kept his Fargo lower than 720 had a mathematically unlikely success rate against players in the 700-720 range in major events it might bear watching. In general though I think people unfairly criticize a handicap system-any of them- because of the potential for abuse IMO. But I am not familiar with all of them of course
 
DeeDeeCues said:
This may seem valid in a perfect vacuum, but Fargo automatically accounts for all of this.

Imagine three people, Al, Bob, and Carl. Al only plays on diamond with Bob. Carl only plays on Valleys with Bob.

Al and Bob play 10000 games and the score is 5000-5000 after all the games. Can we agree that they are equally skilled?

Bob and Carl play 10000 games and again, the score is 5000-5000. Can we also agree they are equally skilled?

So, if Al=Bob and Bob=Carl, I think we would agree that Al=Carl, yet Al and Carl have never played on the same type of table.

You may still argue that Al will beat Carl on dummond tables. I would argue that we have no evidence that is true and have evidence that it is false. Bob is the equalizer. Your logic would say that Bob makes more balls/table runs on the Valley than he does on dummonds. I may even agree that is true, but would point out that he is even with Carl over many thousands of games. We have no reason to assume that Carl's game would suffer any more going to dummonds than Bob's game does.

Now change those three people to groups of hundreds of people and you have described FargoRate.
Click to expand...

ya, but -- there's that 10,00 game thing again.

How is the average sample anything approaching that kind of number?

Lou Figueroa
 
mikepage said:
There is no doubt we can't make meaningful conclusions without lots of games. But look at hist last sentence, Lou, "Now change those three people to groups of hundreds of people..." The key--perhaps counterintuitive key-- here is we don't need lots of games between any two people. The equivalent information to what he is describing is contained in a few games each against many people.
Click to expand...

OK, I think that's partly at the crux of what I've been saying, Mike.

Sample size AND, I feel, "course slope" differentials are also crucial elements that virtually no current system can factor in.

IMO, I think FR is as good as it gets but it is not as perfect as some might present it to be. That's all.

Lou Figueroa
 
Bob Jewett said:
I'll take a stab at it. Here's a hypothetical situation where all of the data is completely valid but the result is crooked.

I have a great table at home and I like to bet high. Five of my friends are pretty close to me and we regularly play races to 5 for $200 to $500. All of the sets are on the up-and-up with everyone trying to win. I send the matches in to FargoRate since we all have ratings and I provide a link to a video of each match so FR can see that all the data is perfectly valid. So far, so good.

The catch is that I only send in the sets where I either lose or win hill-hill. I omit the sets where I dominate. My buddies can't understand why my rating is so far below theirs.

The problem is that the decision to add the data is made after the data is available. That is a huge problem.
Click to expand...

These examples have nothing to do with a match between two top pros not being recorded. Instead of Baseth working towards an established rating he will remain an unknown.
 
Bob Jewett said:
Approach number two: if you have a friend who as their day job works in measurements or statistics, ask them about Mike's restrictions.

I tried.
Click to expand...
Did they answer you? If yes, what was their opinion? If no, never mind, {said in Steve Urkels voice}.
 
You must log in or register to reply here.
Back
Top