Fargo rating

[Woodshaft]

I know you are just speculating. But who are you actually worried about? Did you see our video on league games?

I'm not worried about anyone. It's just that your rating system is based on too many variables to be precise. The games themselves-- 8-Ball, 9-Ball, and 10-Ball, are three totally different animals. Same with table size-- 7 or 9 footers. (Your "Woodward" analysis of table size is meaningless btw. He is an outlier-talent in your data). Or Diamond tables (tight pockets) vs Valley tables (buckets).
Mr. Page, I've been playing for 37 years and have played tons of your "established" fargorated players, both in league play and tournaments. There are, many times, huge differences in players with near identical fargos and their ability to perform equally in each of the three major games or by table size or type. Too many players are one-trick ponies. They excel at one type of game on one type of table, and they play against the same people with the same qualities, and that's how they're fargorated. The data is one-sided. Frankly your "one-number-fits-all-games-and tables", just isn't that accurate in real game situations. A better indicator would be by game type. For example, I personally excel at 8-ball, on any table, and I play even with guys you'd rate about 650. But I hardly play rotation games. Probably 575ish there at best. Why? Because I'm a pure 8-ball player and that's all I play. And I see this discrepancy all the time. Lotsa "one-trick-ponies". Mike, I personally would love to see you do three ratings: A Fargo-8, Fargo-9, and Fargo-10 (ball). You have the data and tools. Those ratings would be MUCH more meaningful, imo.
Cheers!

 

[PoolPlayer4]

Mike Page:

I'd like to see the ability in FargoRate to tell where each player is ranked in the U.S. or world. What would be meaningful to me is to know, am I a top 1000 player in the U.S.? Top 1250? Etc. I'd like to know my exact ranking number and those of other players. I'd also like to be able to tell the distribution of players. How many 700+? How many 650-699? Etc. It seems like these would be easy to do.

Knowing I'm a 650, for example, doesn't tell me that much about where I stand in the sense that I have no idea how many 650s there are in the U.S. or world. The information would be interesting to know and useful in keeping me motivated to improve.

 

[straightline]

There could be parallel divisions - Fargo A, Fargo B, Fargo C etc... that might give a more concise picture of player calibre. Players could also be action rated separately.

 

[mikepage]

Mike Page:

I'd like to see the ability in FargoRate to tell where each player is ranked in the U.S. or world. What would be meaningful to me is to know, am I a top 1000 player in the U.S.? Top 1250? Etc. I'd like to know my exact ranking number and those of other players. I'd also like to be able to tell the distribution of players. How many 700+? How many 650-699? Etc. It seems like these would be easy to do.

Knowing I'm a 650, for example, doesn't tell me that much about where I stand in the sense that I have no idea how many 650s there are in the U.S. or world. The information would be interesting to know and useful in keeping me motivated to improve.

Yes, that's fair.

One thing to consider, though, is we also don't know how many 650s there are in the world. So you would be seeing a number that reflects the level of coverage we have, something that is in flux. There are right now in our system 1,700 players in the us with 100 or more games in the system performing over 650 and 3486 players in the world rated over 650. Those numbers will continue to rise as Fargo Ratings spread.

 

[Bob Jewett]

Yes, that's fair.

One thing to consider, though, is we also don't know how many 650s there are in the world. So you would be seeing a number that reflects the level of coverage we have, something that is in flux. There are right now in our system 1,700 players in the us with 100 or more games in the system performing over 650 and 3486 players in the world rated over 650. Those numbers will continue to rise as Fargo Ratings spread.

Is the shape of the distribution curve changing much? That would at least give him a percentile.

 

[mikepage]

I'm not worried about anyone. It's just that your rating system is based on too many variables to be precise. The games themselves-- 8-Ball, 9-Ball, and 10-Ball, are three totally different animals. Same with table size-- 7 or 9 footers. (Your "Woodward" analysis of table size is meaningless btw. He is an outlier-talent in your data). Or Diamond tables (tight pockets) vs Valley tables (buckets).
Mr. Page, I've been playing for 37 years and have played tons of your "established" fargorated players, both in league play and tournaments. There are, many times, huge differences in players with near identical fargos and their ability to perform equally in each of the three major games or by table size or type. Too many players are one-trick ponies. They excel at one type of game on one type of table, and they play against the same people with the same qualities, and that's how they're fargorated. The data is one-sided. Frankly your "one-number-fits-all-games-and tables", just isn't that accurate in real game situations. A better indicator would be by game type. For example, I personally excel at 8-ball, on any table, and I play even with guys you'd rate about 650. But I hardly play rotation games. Probably 575ish there at best. Why? Because I'm a pure 8-ball player and that's all I play. And I see this discrepancy all the time. Lotsa "one-trick-ponies". Mike, I personally would love to see you do three ratings: A Fargo-8, Fargo-9, and Fargo-10 (ball). You have the data and tools. Those ratings would be MUCH more meaningful, imo.
Cheers!

We hear these ideas all the time, and we see people without names judge themselves as 50 points higher at one game or another or on one table or another. But when these nameless people are converted to actual people with actual games in the system, these effects seem near as we can tell to go away.

If anyone wants to name people who they believe might have many games on different tables or many games playing 8-Ball and rotation for whom they expect a big difference, we'd be happy to investigate.

Note that if someone is basically just a big table player or just an 8-ball bar box player, then the only way to access their Fargo Rating is, respectively, through big-table play or 8-Ball bar box play. That is if they only suck at things they don't play BECAUSE they don't play it, then it it wouldn't be a way to access their Fargo Rating regardless. You don't have a "Fargo Rating" for something you don't play at least a fair amount.

 

[mikepage]

Is the shape of the distribution curve changing much? That would at least give him a percentile.

The shape of the distribution curve has about the same variance (standard deviation close to 100 points) no matter what. It's peak, though, depends sensitively on the minimum number of games in the system.

Amongst 788 people with 3,000 or more games, average rating is 612 and standard deviation is 100
Amongst 6400 people with 1,000 or more games, average rating is 555 and standard deviation is 99
Amongst 16,000 people with 500 or more games, average rating is 526 and standard deviation is 102
Amongst 41,000 people with 200 or more games, average rating is 493 and standard deviation is 108
Amongst 68,000 people with 100 or more games, average rating is 474 and standard deviation is 114

Sure we can say 200 games is what we consider "established," and so the average rating is 493. But we recognize this is a little arbitrary and depends on who is getting games into the system, etc.

 

[PoolPlayer4]

Yes, that's fair.

One thing to consider, though, is we also don't know how many 650s there are in the world. So you would be seeing a number that reflects the level of coverage we have, something that is in flux. There are right now in our system 1,700 players in the us with 100 or more games in the system performing over 650 and 3486 players in the world rated over 650. Those numbers will continue to rise as Fargo Ratings spread.

Thanks for that Mike. How many from 600-649 and 700+, if you have that info handy?

My main point was that this would be a nice built-in feature so a person can see how they are progressing compared to the universe of pool players with Fargo ratings.

 

[The_JV]

I'm not worried about anyone. It's just that your rating system is based on too many variables to be precise. The games themselves-- 8-Ball, 9-Ball, and 10-Ball, are three totally different animals. Same with table size-- 7 or 9 footers. (Your "Woodward" analysis of table size is meaningless btw. He is an outlier-talent in your data). Or Diamond tables (tight pockets) vs Valley tables (buckets).
Mr. Page, I've been playing for 37 years and have played tons of your "established" fargorated players, both in league play and tournaments. There are, many times, huge differences in players with near identical fargos and their ability to perform equally in each of the three major games or by table size or type. Too many players are one-trick ponies. They excel at one type of game on one type of table, and they play against the same people with the same qualities, and that's how they're fargorated. The data is one-sided. Frankly your "one-number-fits-all-games-and tables", just isn't that accurate in real game situations. A better indicator would be by game type. For example, I personally excel at 8-ball, on any table, and I play even with guys you'd rate about 650. But I hardly play rotation games. Probably 575ish there at best. Why? Because I'm a pure 8-ball player and that's all I play. And I see this discrepancy all the time. Lotsa "one-trick-ponies". Mike, I personally would love to see you do three ratings: A Fargo-8, Fargo-9, and Fargo-10 (ball). You have the data and tools. Those ratings would be MUCH more meaningful, imo.
Cheers!

Are you sure you didn't mean that "isn't" based on enough variables to be precise...? The rest of your post calls out the use of a blanket system that group players with varying skill sets...

 

[The_JV]

Though for simplicity I would probably create a Fargo specific scale to rather than trying to apply existing scales to Fargo. For example, 400-499 is a C player, 500-599 is a B player and so on. Then for 700+ players, 700-749 is a semi-pro, 750-800 is pro and 800+ is world class. That may not align with existing rating scales, but it’s easy to remember.

..but doesn't a difference of ~100 points in fargo translate to a player being 2x better than the other...? Thought I read that somewhere. If it is indeed true. Don't you think the single letter designation for a 400 player vs 499 player misleading...? At the least I think you'd need to split that difference by at least one more increment.

 

[The_JV]

Fargo Ratings to them seem like using metric units for speed or distance (km/hour or meters) and they just are not yet used to thinking in terms of them. It is not a good time to let this tail way this dog, imo.

Nailed it... drop mic... end discussion...

There's a precision with Fargo rate that has to be admired. Yes players can play lower or higher than their rating at any moment, but there is a definitive difference between a 650 and a 680 player. Furthermore, and what I love the most about the system, is the competition it can instill between evenly matched players. There's a local player with a 9 point advantage on me right now. Sure my fargo isn't established as of yet, but that doesn't change the fact that I'm pumped to try and squeek ahead of him by the time my robustness becomes valid. In the world of letters we'd both just be 'A's.... boring.

 

[jalapus logan]

If you send an email to support@fargorate.com, we can get the games assigned to the correct records
Thank you, I sent an email to that address. Happy new year.

 

[Bob Jewett]

..but doesn't a difference of ~100 points in fargo translate to a player being 2x better than the other...? Thought I read that somewhere. If it is indeed true. Don't you think the single letter designation for a 400 player vs 499 player misleading...? At the least I think you'd need to split that difference by at least one more increment.

There are lots of details about the system in the FAQ list on the FargoRate site.

Briefly, if two players are 100 points apart the better player will win in a 2:1 ratio. If they play 12 games, the most likely score is 8-4.

If two players are 200 points apart, the better player will win in a 4:1 ratio. 20 games is expected to finish around 16-4.

If two players are 300 ... 8:1

400 ... 16:1

The top players are over 800 and there are lots of 400 players in the system. Even 100-rated players exist. A 400 beats a 100 like a rented mule but might not get 1 game in a race to 20 against Shane. The range of playing ability at pool is remarkable.

As for the brokenness of assigning a few letters to a range of ratings and potentially having players one rating point apart ... -- well, yes it is horribly broken. That's why all systems that use a few rating levels are broken. Another good example is a league that rates players from 2 to 9. Such systems are not trying to be accurate.

 

[Cameron Smith]

..but doesn't a difference of ~100 points in fargo translate to a player being 2x better than the other...? Thought I read that somewhere. If it is indeed true. Don't you think the single letter designation for a 400 player vs 499 player misleading...? At the least I think you'd need to split that difference by at least one more increment.

Sure, those were just examples. Increments of 50 make more sense. It’s still a meaningful difference, but the player at the lower end still has a decent chance.

 

[Bob Jewett]

Sure, those were just examples. Increments of 50 make more sense. It’s still a meaningful difference, but the player at the lower end still has a decent chance.

For two players 50 points apart...
race to 11, 79%-21%
race to 5, 70%-30%

If they are 100 points apart, the weaker player is 1 in 20 in a race to 11.

 

[Cameron Smith]

I'm not familiar with how people communicate in the chess world. I'm curious, though. I know there are various A/B/C/D classes with rating ranges that people use for tournament classes within certain organizations. We are working with a country federation that will have tournament classes based on Fargo Rating ranges. But another bigger or smaller country might find different ranges useful. I don't see some universal impetus to define a class that has universal meaning. Attached is a chart that shows some overlapping classifications. In a bizarre coincidence, it is prepared by Chess's "Dr. Dave."

With that said, though, we DO need to have ways to express a player-level qualitatively. Defining A,B,C, D, E based on Fargo Rating ranges is one way to do that. But I have two concerns.

The first is if we use letter designations that already have meaning to people (like B and A), we are inviting confusion. To someone in poland, a "B"might be someone who can't quite hang with a world-class player, and to someone in Oregon, a B might be someone who runs a table once in a blue moon when all the stars align. If you want to fix that communication problem by assigning a universal rating range, then why not create new labels that don't have baggage?

The second is we should be looking at desire paths. The country of Finland sends people out after a new snowfall (that covers up paved paths) to see where people actually walk. What path across a field do they take that might be an efficient way to get from the subway exit to a particular street and have a bit of nice lake view? What paths do students on a campus actually take between the lecture halls and the cafeteria and the dorms and the parking lots. Some colleges choose to pave no paths until they see where the grass is worn. You are kind of crowd sourcing the design like this.

In a similar way, we look to places like Oklahoma and Arizona and Wisconsin--places where everybody knows Fargo Ratings and they are used all the time. This is where everyplace is heading. What do they actually do when they want to convey qualitatively how someone plays? Oklahoma and Arizona were both entrenched five years ago with numbers. An Arizona "7" was centered around about 500, and an Oklahoma "7" was maybe 40 points higher. They used to talk about 6's and 7's and 9's, etc all the time. Do they still find numbers like that useful to represent a qualitative idea or a range? Actually no. Check out the facebook groups. There are no desire paths involving these numbers in those places. If people want to communicate speed qualitatively, they will use just one significant figure (500 level, 600 level, 700 level, etc). If they want to refine a bit, it's 400, 450, 500, 550, etc. And this works fine.

The impetus to translate from Fargo Ratings to some old local ratings just comes from people not yet accustomed to Fargo Ratings. So they can't imagine. Fargo Ratings to them seem like using metric units for speed or distance (km/hour or meters) and they just are not yet used to thinking in terms of them. It is not a good time to let this tail way this dog, imo.View attachment 579902

Sure, I can see confusion using the letter ratings. I only suggested that because they appear to be common place in other sports. Though, the saving grace would be that at least they would be calibrated to Fargo. Prior to Fargo, people may as well have been just speaking different languages.

I understand the impetus to move towards only communicating ratings using the numbers, and that might be the future. But the skill classes help contextualize the numbers in a consistent way and this is particularly important for new players. I don’t have a lot of experience in the chess space, but I do mostly hear people communicating with ELO numbers, but the class system helped me and others understand what those numbers mean.

So for me, some sort of category breakdown that is easy to communicate just helps players learn the ratings and potentially accelerate their ability to communicate skill levels using the numbers.

 

[hang-the-9]

There are no universal uses for the letters. There is no "true B speed." Poland does handicapped tournaments in which a "B" is around 680 and a "C" might be 650. In Minnesota a typical "B" is in the upper 400s and an "A" is maybe 560 give of take 30. points. US mid-atlantic/northeast you'll see various things between these. In Michigan there is a BB rating in the upper 500s or so. These letters just need to go away.

What's a B bowler? What's a B golfer? What is "B level" for the pole vault or the mile run? What is a "B level" chess player?

The letters just need to go away.

But this is not an issue with the rating system, it's an issue with people not using it correctly. ABCD, A is strong, D is a bit over beginner. Thus a C has no business being a 650 unless the people using this rating are clueless where it comes from. Like I said in another post, it's logic and reason. You can't have a C player running racks if a D is a beginner sine C is just one level over D, and you can't have a B player running racks since that would mean a A has to be Pro level, which is silly since there are a few levels over A. That would make a Pro never missing a shot in 3 years of playing. It all comes together pretty easily if people are not stubborn and ignore reason. Start at the two ends and use some brain to fill in the slots in the middle.

 

[hang-the-9]

That is the nature of local ratings. It is not uncommon to call the best players "A," the first letter in the alphabet. Then call the next group down "B," and the next "C," and so forth. That is the point here with the Poland group. The top level is world top 100 speed players. So you might call them "A." Then lesser elite players might be called "B," and so forth.

Or you can just forget the letters.

In the ABCD rating the As are common good players, there is also an Open and then Pro level over that, and the A has A- and A+ and maybe even go into A++, although to me that is Open, and then the A+++ are the Pro level players that are bellow world beater class. It's a lot better than the league ratings, where they compare relative skill in a small group of local players, the ABCD rating measures your knowledge in general (or at least that is what the correct way of using it is). How good can a D run through a rack vs how good can a A run through a rack before they miss, how well do they know safety shots and how well they execute that, do they know position play? Really just 3 or 4 factors to get the skill rating based on how much they know and how well the execute.

 

[JimJones1]

I don't really understand Fargo, It doesn't make much sense to me at all, because when we are all in the pool room nobody I know talks about it. What tournaments are happening where they ask people of specific ratings they cant play, this is the purpose of Open Tournaments. Or you can play a tournament that has a handicap system. I like the (D, C, B, A, AA, AAA, Open, & Pro) rating more because we all know about what speeds those are; albeit, most people think they play above the speed they actually do.

 

[MattPoland]

I don't really understand Fargo, It doesn't make much sense to me at all, because when we are all in the pool room nobody I know talks about it. What tournaments are happening where they ask people of specific ratings they cant play, this is the purpose of Open Tournaments. Or you can play a tournament that has a handicap system. I like the (D, C, B, A, AA, AAA, Open, & Pro) rating more because we all know about what speeds those are; albeit, most people think they play above the speed they actually do.

In my experience most people pretend they are a lower speed than they actually are to steal a tournament. And there’s a lot of drama and politics around who gets bumped up and when. Usually when a bracket gets to the Calcutta, you’re looking at all the players that should never have been in the bracket. It’s not always fair for some of the people at the lower end of the bracket because they end up fodder.

I’ve also seen how drastically different ABCD can be from state to state. Here’s how Wisconsin ratings compare to Michigan ratings...

WI GM = MI AAA
WI M = MI A and AA
WI AA = MI BB and half of B+
WI A = MI our B+ and our B
WI B = MI C and C+
WI C = MI D and D+

In Michigan we are still seeing ABCD tournaments but they are starting to use FargoRate to as the primary cutoff for which bracket you’re in. Like this flyer for a BB and under tournament.

Sent from my iPhone using Tapatalk Pro