A Questoin For FargoRate

[watchez]

If Justin Bergman beats

1. Darren Appleton in a race to 5 in the Mosconi Cup
   
2. Darren Appleton in a race to 100 for $20,000
   
3. Darren Appleton in a race to 11 winner breaks in the US Open
   
4. Darren Appleton in a weekend tourney, alternate breaks on a bar box race to 7

Do they all count the same?

 

[BRussell]

Maybe Mike Page will answer, but here's my guess:

Games in the system are all equal (except that they decrease in weight over time). So it's not the matches that are equivalent, like winning a race to 5 = winning a race to 100, it's the games within those matches that count. Winning 100 games will affect the rating more than winning 5 games.

But I believe each game is equivalent, bar box, Mosconi Cup, private match, etc., as long as they go into Fargo of course. I'm sure they could weight them differently, like they do for time, but it would be subjective.

 

[skip100]

Hopefully there is a way to distinguish between sessions, or at least have diminishing returns with a very long match. I think most of us would agree that one race to 50 should not be worth the same as five races to 10.

 

[Jeff G. Martin]

I can't speak for FargoRate (but I'm sure Mike or Steve will chime in), but a similar ELO-based performance rating system is weighted as follows:

Tournament = 1.0
Informal = 0.8
League = 0.6
Minis = 0.6

"Informal" being a typical gambling session. I'm not so sure a match for $20,000 is "typical", though, and as such might carry more weight.

Regarding winner break formats: I wonder how this is handled in the algorithm. If FargoRate says an 800 should beat a 700 2:1, this is assuming both players get to shoot at least every other game, right? I'm only assuming.

 

[BmoreMoney]

Just my opinion, but I do whole heartedly believe a gambling session deserves to be weighted the most heavily. Show and reward true heart.

 

[JB Cases]

Everything has to do with who else they play and beat. Justin could win against Darren in all those situations but then Darren could go and beat players like Ko who in turn beat more players of Darren and Ko's caliber throughout the year....essentially everyone is tied to everyone else's performance.

 

[Saturated Fats]

...one race to 50 should not be worth the same as five races to 10.

Maybe, but I don't see why.

 

[skip100]

Maybe, but I don't see why.

Unlike, say, chess, which is well suited for a pure ELO system, pool players can have huge variations in the quality of their play from day to day. Becoming a good player is not just about how high your highs are, but how low your lows are.

Imagine the scenario where player B beats player A in five consecutive tournaments by an average score of 10-8. Then A beats B 50-39 in a one-night challenge match. The overall record would then be 80 wins for A and 79 wins for B. What's more likely - player A is better than player B, or player B has A's number but had a bad night in the challenge match? I think all of us would be betting on player B to win the next match in that situation even though he has a losing record.

 

[mikepage]

If Justin Bergman beats

1. Darren Appleton in a race to 5 in the Mosconi Cup
   
2. Darren Appleton in a race to 100 for $20,000
   
3. Darren Appleton in a race to 11 winner breaks in the US Open
   
4. Darren Appleton in a weekend tourney, alternate breaks on a bar box race to 7

Do they all count the same?

First, congratulations to Justin on that fine series of performances...

BRussell is right. Other than the time decay factor individual games are given the same weight. This likely won't always be the case. Basically there is a tradoff between refined information and statistically significance. Right now that balance is shifted in the direction of needing statistical overall significance.

Here is SVB's record against a couple opponents:

Against Corey:

22 to 19 on 7' table (53.6%)
95 to 83 on 9' table (53.4%)

Against Jeremy Sossei

24 to 17 on 7' table (58.5%)
38 to 25 on 9' table (60.3%)

These numbers are kind of small, but it is looking from this like Corey is a little better player than Jeremy and that Shane is better than both. We can more or less draw at least tentatively that same conclusion from just the 7-foot results or just the 9-foot results.

But with the results combined we gain more confidence. That confidence comes at the expense of washing out the real subtle differences between how these players play on 7-foot or 9-foot tables. Those subtle differences are small though, and for now we have good reason to ignore all the implied distinctions in the question.

 

[Teacherman]

This is no more accurate than the APA nonsense.

Garbage in....garbage out.

When Justin played Jayson Shaw, Jayson quit twice. He quit the first day after Justin got some rolls and put a barage on him. Then after Justin stymied his comback on day 2....he quit again.

In their race to 80, I'd say at least 30 games were not legit.

 

[Nostroke]

It's a good system that will only get better-It will never be perfect and include every possible nuance. It would be sad if it did get perfect. The players wouldnt even have to show up for a tournament. Just do the draw, advance the players by their rating and Presto!!! We have our winner!!

 

[BRussell]

This is no more accurate than the APA nonsense.

Garbage in....garbage out.

When Justin played Jayson Shaw, Jayson quit twice. He quit the first day after Justin got some rolls and put a barage on him. Then after Justin stymied his comback on day 2....he quit again.

In their race to 80, I'd say at least 30 games were not legit.

You mean he forfeited all those games? If so, that's right, it shouldn't go into the rating.

 

[JB Cases]

It's a good system that will only get better-It will never be perfect and include every possible nuance. It would be sad if it did get perfect. The players wouldnt even have to show up for a tournament. Just do the draw, advance the players by their rating and Presto!!! We have our winner!!

It could never be that way because of the ever changing skill sets of the players. Players are always getting better and worse for a variety of reasons.

 

[robsnotes4u]

This is no more accurate than the APA nonsense.



Garbage in....garbage out.



When Justin played Jayson Shaw, Jayson quit twice. He quit the first day after Justin got some rolls and put a barage on him. Then after Justin stymied his comback on day 2....he quit again.



In their race to 80, I'd say at least 30 games were not legit.

Hmm, and I thought you wanted to help promote, and grow the game of pool.

I would guess 30 games don't mean a lot when your robustness is almost 3000.

May I suggest you go to www.fargorate.com, look at the website and watch the videos? It just might give you a better understanding of what FargoRate does and why, that is if you are truly trying to promote, and grow pool.


Thanks, and have an awesome day





Sent from my iPad using Tapatalk

 

[watchez]

I went to the website. Didn't watch the videos. I can tell you between the background, the font used, the color of the font and basically the organization of the site itself it left me more lost than it did before I entered the site. Not trying to be negative but you brought up the site. That's why I posted my question here.


And I have read Mike's response here three times and it still isn't clear to my questions. Specifically the last two sentences of his second paragraph. Maybe he is being cloudy cause he doesn't want to give away his secret of the sauce so to speak and have someone immitate it. Or maybe my question is valid and wasn't thought of when Fargorate was created.

 

[iusedtoberich]

Mike, why don't you explicitly show the formula used to compute the ratings? I think that would clear up a lot of questions on how the system works. If you do, please do it on your site (and link here) so that it is there for everyone to see. Thanks.

 

[robsnotes4u]

I went to the website. Didn't watch the videos. I can tell you between the background, the font used, the color of the font and basically the organization of the site itself it left me more lost than it did before I entered the site. Not trying to be negative but you brought up the site. That's why I posted my question here.


And I have read Mike's response here three times and it still isn't clear to my questions. Specifically the last two sentences of his second paragraph. Maybe he is being cloudy cause he doesn't want to give away his secret of the sauce so to speak and have someone immitate it. Or maybe my question is valid and wasn't thought of when Fargorate was created.

I can not speak for Mike, and it is very clear to me as i have had the luxury of discussing, and playing within the Fargo Rate for 5 years. With that, with his scientific, mathematical, and research background I doubt any stone has been left unturned.

Sorry that your perception of the site is different than mine. Nothing I can do about that.

I hope the best for you in finding the answers to your questions.





Sent from my iPhone using Tapatalk

 

[railbird99]

This is no more accurate than the APA nonsense.

Garbage in....garbage out.

When Justin played Jayson Shaw, Jayson quit twice. He quit the first day after Justin got some rolls and put a barage on him. Then after Justin stymied his comback on day 2....he quit again.

In their race to 80, I'd say at least 30 games were not legit.

So just to clarify, are you saying the rating system itself isn't accurate because it can't turn garbage into gold, or that the garbage (being the matches) are equivalent to "APA nonsense"?

 

[railbird99]

Mike, why don't you explicitly show the formula used to compute the ratings? I think that would clear up a lot of questions on how the system works. If you do, please do it on your site (and link here) so that it is there for everyone to see. Thanks.

Lmao. So it's widely known that it's ELO ratings, and even with all of the information and simple examples at your disposal online, you are still baffled by it, but somehow reading an algorithm is going to alleviate all of the confusion?

He's not hiding it as a method of obfuscation. There are variations of the ELO system, and tweaking that can be done to improve upon it. He's obviously put a lot of work into this, and doesn't want someone just copying it.

 

[railbird99]

I went to the website. Didn't watch the videos. I can tell you between the background, the font used, the color of the font and basically the organization of the site itself it left me more lost than it did before I entered the site. Not trying to be negative but you brought up the site. That's why I posted my question here.


And I have read Mike's response here three times and it still isn't clear to my questions. Specifically the last two sentences of his second paragraph. Maybe he is being cloudy cause he doesn't want to give away his secret of the sauce so to speak and have someone immitate it. Or maybe my question is valid and wasn't thought of when Fargorate was created.

It really boggles my mind why people seem so confused and lost with this. Fargo is not the first ELO-based rating system, and pool isn't the first sport to use it. The internet has all the information you could ever want on the subject, but instead of trying to understand the fundamentals of what Fargo is based on first, you try to poke holes in something you know nothing about.

I'm pretty sure your original question was answered as clear as possible. It's each individual game that counts, not the overall match, and each individual game is treated the same, regardless of how long the race is, what kind of match, game rules, or what size table, etc.

If you would like to understand the significance of that, and why it is the way it is, google is your friend.