Fargo Rating? Valley vs Diamond

So you might recommend that when playing on a 7-footer, people should take the closer of the possible spots that Fargo recommends?

No. That doesn't matter. Those are close to 50/50 either way. But if you choose to be in a game for which you are at a disadvantage, Shorter sets and smaller tables are your friend. And odds on the $$ is your best friend;-)
 
Yes, and Fargo Ratings don't need to take it into account.

Here's why.

Rory H, my house pro, and I both have about 6,000 games in the system, including hundreds against each other on both 7' and 9' tables.

We are 102 points apart, so he is supposed to win games over me in a 2 to 1 ratio, and he does. In fact he's 2 to 1 over me on 9-foot tables, 2 to 1 over me on 7' tables, 2-to-1 over me in rotation games, 2-to 1 over me in 8-ball. We've also played plenty of straight pool, and--you guessed it--he's 2-to-1 over me.

If we play an even-up race to 7 in a tournament, he is a strong favorite over me.

I am way more likely to beat him in a race to 7 on a 7-foot table than on a 9-foot table. You may say, then aren't I closer to him in Fargo Rating on a 7-foot table than a 9-foot table? It may seem that way, but the answer is no. We are still 100 points apart.

So what's going on?

The easier table, where I might run 2 racks out of 7 and he might run 4 racks out of 7, makes a race to 7 effectively a shorter race. So it statistically acts more like a race to 5 does on a 9-foot table, where I'm more likely to catch a fluctuation and win the set.

An extreme example of this is straight pool. Bob Jewett recently beat Shane Van Boening in a race to 125. I am confident if they raced to 10,000, Bob would not reach 5,000; that is I am confident Bob is more than 100 points below Shane. Yet he won a race to 125. Could you imagine Bob winning a 9-ball race to 125? I didn't think so. The difference is a race to 125 in straight pool is fewer innings, fewer changes in control. It is probably more like a race to 7 or 9 in 9-ball.

MIke, speaking of straight pool, do you feel that the current Fargo Ratings fairly represent straight pool strength for players currently in the system even though, as I understand it, straight pool games are not logged into the FargoRate system?

Thanks,
Will
 
MIke, speaking of straight pool, do you feel that the current Fargo Ratings fairly represent straight pool strength for players currently in the system even though, as I understand it, straight pool games are not logged into the FargoRate system?

Thanks,
Will

I do, with the caveat that the players have at least some modest experience playing straight pool
 
No. That doesn't matter. Those are close to 50/50 either way. But if you choose to be in a game for which you are at a disadvantage, Shorter sets and smaller tables are your friend. And odds on the $$ is your best friend;-)

It can't both be true that 1) playing on a 7-footer is like playing a shorter race and 2) you don't have to adjust FargoRate predictions when playing on a 7-footer.
 
The difference between Mike and others on here is simple to me. Mike is able to see the results of every game in the system. The individuals on here can only draw conclusions on a few games of an individual. Mike's approach removes any preconceived notions. Mike's post regarding the play between him and Rory is a great example. I have played them both a lot in every game besides one pocket, on different tables. No matter what my thought is on each player, and what I have seen, my perception is not as accurate as FargoRate. This doesn't change your perception, as you don't have the same experience


Sent from my iPad using Tapatalk
 
How exactly do these matches end up in the system? If there are millions of matches in there, how many sources are entering them?
 
It can't both be true that 1) playing on a 7-footer is like playing a shorter race and 2) you don't have to adjust FargoRate predictions when playing on a 7-footer.

It depends on what you mean by "adjust the FargoRate predictions"

--ratings and ratings differences: stay the same
--fair matches: stay the same
--chance of each player winning an unfair match: changes and is consistent with the shorter, effective race length.
 
FargoRate is not perfect. FargoRate will never be perfect. FargoRate can never be perfect. And the reason is because no system can ever be perfect. What happens is that at a certain point as you start getting close to perfection, anything you can “improve” causes another thing to be not as good. So once you get to that point it just becomes a compromise or balancing act on what is best overall (with all things considered), and then this is literally as good as it can get. No other system has ever come even remotely close to getting to this point, but FargoRate has gotten there IMO.

With this in mind, that it is literally impossible for any system to ever be perfect because of the trade offs that have to be made, here are what I believe are the only important questions to ask ourselves regarding FargoRate:
-Is it better than not having it at all?
-Is it better than anything else that is trying to do the same thing?
-Can it be improved in any meaningful and practical way for the purposes and uses for which it is intended? Or to rephrase that a different way, for the purposes of being able to rate and compare all pool players in the world against any other pool players in the world, are there any improvements that can be made that would not come at the cost of something that is equally or more detrimental?

I say the answers are yes, yes, and no.

Is it better than not having it at all?
I say yes if it meets two caveats, which are that it be at least reasonably good and not wildly inaccurate, and that it be better than anything that already exists for the same purpose. I think pretty much everybody agrees that it meets those two criteria, and therefore it is better than not having it at all. Nobody thinks it wildly inaccurate, and everybody agrees that it is the best system for rating and comparing any two pool players in the world because it is the only system that lets you rate and compare any two pool players in the world. Even when compared to other rating systems that are designed to compare players only in a smaller area like locally or nationally, I think everyone agrees that it still even does a better job there too. So no question, it is worth having.

Is it better than anything else that is trying to do the same thing?
No point in having something if something else before it already does a better job right? As was already covered above, there is no other rating system that can or is even trying to do the same thing, which is to be able to rate and compare players worldwide in a consistent way. Clearly FargoRate is better than what already exists for this purpose because nothing else exists for this purpose. And even for those systems only intended to rate players in a smaller group or area FargoRate clearly does a better job there too. There doesn’t seem to be any dispute about this (I sure haven’t seen any anyway) but if for some reason you believe the APA or any other rating system out there is more accurate feel free to humiliate yourself by giving your argument.

-Can it be improved in any meaningful and practical way for the purposes and uses for which it is intended? Or to rephrase that a different way, for the purposes of being able to rate and compare all players in the world against any other players in the world, are there any improvements that can be made that would not come at the cost of something else that would be at least equally detrimental?
No. Could the system by made even more accurate if it were only used for professional players? A little bit. Could the system be made even more accurate if it were only used for the game of 8 ball and nothing else? A little bit. Could the system be made even more accurate if it were only used for 9 foot Diamond tables with factory cut 4.5” pockets with brand new properly installed simonis 860 cloth when the humidity is exactly 10%, etc? It could.

But when you start doing these or any other “improvements” it comes with an even bigger drawback. If you limit it to pros, then it doesn’t help and isn’t usable by the other 99.99% of the rest of the world. If you limit it to only 8 ball then you severely limit the amount of data and sample size and even exclude altogether the players who don’t play that game. If you limit it to certain types of equipment you have the same issues.

And if you try to account for all the possible variables in the formula such as the table size, cloth type and condition, pocket size, pocket cut angles, shelf depth, humidity levels, barometric pressure, how drunk or sober each opponent was, how sick or well they each were, how in stroke or out of stroke they each were, etc, etc, etc, etc, etc, the formula becomes too complicated, and there would be way too much data to input. Mike would need a team of 50 data entry people to keep up with it even if the formula could accurately take all these things into consideration.

I’m also pretty certain it is hard enough to get league operators and tournament directors to report all their match statistics as it is right now when all they have to report for each match is the identity of the players, who won, and the scores. Many people are inherently lazy, especially if they don’t feel that something is directly benefitting them right now at this moment. Can you imagine if they had to also report all table conditions, or the weather conditions, or the conditions of the opponents (drunk, in stroke, getting over the flu) too? Forget about it, nobody would ever submit any data to Mike if it got much more involved than the players and the score. Not to mention that much of that other stuff is totally subjective anyway and whatever they reported wouldn’t even be reliable. These are just a few of the many examples of how “improving” any one thing is going to make something else much worse.

The fact of the matter is that FargoRate is about as perfect as is possible for something which is intended to be able to rate and compare all the pool players of the world against each other. Any little thing you could make a little better in one way just makes something else much worse. Right now it is about the perfect all around compromise as it is, and the nice thing, should Mike ever see the need for it and want to have a second or third “version”, is that FargoRate could be tweaked to be even a little bit better for any very specific application, such as for 8 ball only, or if used just for the pros.

The one (and really only) thing that can make it better right now as a system for every pool player in the world, and which everyone already universally agrees on, is more data. The more match statistics there are for everyone, the better and more accurate it gets (and it is already pretty darn good). So the best thing, and really the only thing that can be done to improve it at this point is for all of us to ask every league and tournament director to submit all their match statistics to Mike for inclusion.
 
Last edited:
^^^^^^^^^^^

That was a very thorough and well written comment above. I enjoyed the input.

If I had my wish list, #1 would be to factor Dr. Dave's TDF into the match as it's entered into the Fargo Rate system. I still believe 7-foot table results should be weighted differently than the big tables-or more specific matchup tables (like the ones at hard times and Steinway with less than 4" pockets).

I don't play golf but I understand there is a central computer system that an amateur enters his results in to prove his handicap. I also understand the particular course this amateur just played has a difficulty rating that is factored in. So it makes sense to use the tables difficulty rating in pool.
 
If I had my wish list, #1 would be to factor Dr. Dave's TDF into the match as it's entered into the Fargo Rate system.

I love Dr. Dave's table rating system. Here are what I see as the problems with incorporating it into the FargoRate formula though:

Measuring accuracy or trustworthiness. Have you seen the way some people measure things? I don't know if people could or would be accurate enough in measuring things for it to add a level of accuracy to FargoRate. And that is assuming they actually tried and didn't just "guesstimate" because they didn't have a ruler handy or were just plain lazy which obviously adds a whole new dimension to whether those numbers could be trusted or not.

Diminished returns. I think table variety issues largely even themselves out and come out a wash in the long run. If there was an improvement to the FargoRate ratings by having the pocket information (and I see no point in having table size if you aren't going to include pocket information too), it would likely be small and I think there is some real question if it would be an improvement at all because of the inaccuracies that will result for the reasons listed above.

Data entry for Mike. At a certain point it isn't worth the time, effort and expense it takes for Mike to do all this data entry. I don't know exactly where that point is, but I can't imagine he is will be retiring and buying a 150 foot yacht off of this. It seems to be a labor of love more than anything and with too much data entry I could easily see it becoming a labor of hate that is no longer worth the hassle.

And perhaps the most important reason is that I think if you add much if anything else to what people have to submit to Mike they just aren't going to bother, and then there goes the most important component to the success and accuracy of the system--getting the data, and as much as possible.
 
I love Dr. Dave's table rating system. Here are what I see as the problems with incorporating it into the FargoRate formula though:

Measuring accuracy or trustworthiness. Have you seen the way some people measure things? I don't know if people could or would be accurate enough in measuring things for it to add a level of accuracy to FargoRate. And that is assuming they actually tried and didn't just "guesstimate" because they didn't have a ruler handy or were just plain lazy which obviously adds a whole new dimension to whether those numbers could be trusted or not.

Diminished returns. I think table variety issues largely even themselves out and come out a wash in the long run. If there was an improvement to the FargoRate ratings by having the pocket information (and I see no point in having table size if you aren't going to include pocket information too), it would likely be small and I think there is some real question if it would be an improvement at all because of the inaccuracies that will result for the reasons listed above.

Data entry for Mike. At a certain point it isn't worth the time, effort and expense it takes for Mike to do all this data entry. I don't know exactly where that point is, but I can't imagine he is will be retiring and buying a 150 foot yacht off of this. It seems to be a labor of love more than anything and with too much data entry I could easily see it becoming a labor of hate that is no longer worth the hassle.

And perhaps the most important reason is that I think if you add much if anything else to what people have to submit to Mike they just aren't going to bother, and then there goes the most important component to the success and accuracy of the system--getting the data, and as much as possible.

This is all valid and I absolutely understand.
However, would it be possible to have the data entered just once and carry over?
So if Steinway registers and gives all TDF data and table numbers, then when a match is played on table 3 with the resulting score- the system already has table 3 TDF entered so no need for extra data from the TD or data entry.

This would go the same for all Pool halls wherever a tournament is held. To qualify for Fargo Rate, they would simply have to register the table numbers and TDF data, and not touch it after that...unless of course a table gets modified or replaced.

And of its a major tournament with standard diamonds tables, 9 foot 4.5" corners like TS or DCC, then that data is already entered because each table is the same. He could also have a stock TDF for un modified valley tables too for tournaments held in bars.

I know it's wishful thinking, but it would really make the Fargo Rate that much more advanced.
 
Where do you stop; age of cloth, type of cloth, temperature of table, humidity, etc?

I talked to Steve Ernst at Fargo Billiards on Friday night and they love to take the data and test everything. That is fun for these guys. If any of it made a big difference they would do something with it. Mike and Steve are special individuals that do this because they care and love it


Sent from my iPhone using Tapatalk
 
Where do you stop; age of cloth, type of cloth, temperature of table, humidity, etc?

I talked to Steve Ernst at Fargo Billiards on Friday night and they love to take the data and test everything. That is fun for these guys. If any of it made a big difference they would do something with it. Mike and Steve are special individuals that do this because they care and love it


Sent from my iPhone using Tapatalk

I fully agree. I don't know Steve but I do know Mike and I trust that he wouldn't be a part of anything like this if he hadn't fully vetted it as much as humanly and scientifically possible.

I am not sure that 200 games is enough to really call a rating established as I see some players who are established who play even with me and better than me who have lower ratings and one who isn't as good as me who is 35 points higher.

And to be clear, IF we all matched up using Fargo's fair match predictions I do not think I could out run the spots I am supposed to give to the lower ranked players on my list and the higher ranked one wouldn't dare try to spot me what Fargo says is a fair game.

So I personally wonder if there isn't a higher number which is going to be more accurate. Maybe my number is actually too low because I only have 250 games in. The guy ranked above me who is even with me at best has 1200 games in. So I have to feel that his ranking is where it should be.

The guys below me who have lower rankings but who are even or better have between 350 and 250 games in. So they are ranked too low based on known ability for me to gamble with them under fargo's predictions. Nor could I give them any spot in a tournament and have a good chance to go hill-hill as fargo predicts a match should go when players are evenly matched according to their rating.

Overall I am heavily pro-Fargo ratings. Personally though it's tough to fade the anomalies if people ask to use Fargo to match up but known ability clearly belies what the ratings say.
 
FargoRate is not perfect. FargoRate will never be perfect. FargoRate can never be perfect. And the reason is because no system can ever be perfect. What happens is that at a certain point as you start getting close to perfection, anything you can “improve” causes another thing to be not as good. So once you get to that point it just becomes a compromise or balancing act on what is best overall (with all things considered), and then this is literally as good as it can get. No other system has ever come even remotely close to getting to this point, but FargoRate has gotten there IMO.

With this in mind, that it is literally impossible for any system to ever be perfect because of the trade offs that have to be made, here are what I believe are the only important questions to ask ourselves regarding FargoRate:
-Is it better than not having it at all?
-Is it better than anything else that is trying to do the same thing?
-Can it be improved in any meaningful and practical way for the purposes and uses for which it is intended? Or to rephrase that a different way, for the purposes of being able to rate and compare all pool players in the world against any other pool players in the world, are there any improvements that can be made that would not come at the cost of something that is equally or more detrimental?

I say the answers are yes, yes, and no.

Is it better than not having it at all?
I say yes if it meets two caveats, which are that it be at least reasonably good and not wildly inaccurate, and that it be better than anything that already exists for the same purpose. I think pretty much everybody agrees that it meets those two criteria, and therefore it is better than not having it at all. Nobody thinks it wildly inaccurate, and everybody agrees that it is the best system for rating and comparing any two pool players in the world because it is the only system that lets you rate and compare any two pool players in the world. Even when compared to other rating systems that are designed to compare players only in a smaller area like locally or nationally, I think everyone agrees that it still even does a better job there too. So no question, it is worth having.

Is it better than anything else that is trying to do the same thing?
No point in having something if something else before it already does a better job right? As was already covered above, there is no other rating system that can or is even trying to do the same thing, which is to be able to rate and compare players worldwide in a consistent way. Clearly FargoRate is better than what already exists for this purpose because nothing else exists for this purpose. And even for those systems only intended to rate players in a smaller group or area FargoRate clearly does a better job there too. There doesn’t seem to be any dispute about this (I sure haven’t seen any anyway) but if for some reason you believe the APA or any other rating system out there is more accurate feel free to humiliate yourself by giving your argument.

-Can it be improved in any meaningful and practical way for the purposes and uses for which it is intended? Or to rephrase that a different way, for the purposes of being able to rate and compare all players in the world against any other players in the world, are there any improvements that can be made that would not come at the cost of something else that would be at least equally detrimental?
No. Could the system by made even more accurate if it were only used for professional players? A little bit. Could the system be made even more accurate if it were only used for the game of 8 ball and nothing else? A little bit. Could the system be made even more accurate if it were only used for 9 foot Diamond tables with factory cut 4.5” pockets with brand new properly installed simonis 860 cloth when the humidity is exactly 10%, etc? It could.

But when you start doing these or any other “improvements” it comes with an even bigger drawback. If you limit it to pros, then it doesn’t help and isn’t usable by the other 99.99% of the rest of the world. If you limit it to only 8 ball then you severely limit the amount of data and sample size and even exclude altogether the players who don’t play that game. If you limit it to certain types of equipment you have the same issues.

And if you try to account for all the possible variables in the formula such as the table size, cloth type and condition, pocket size, pocket cut angles, shelf depth, humidity levels, barometric pressure, how drunk or sober each opponent was, how sick or well they each were, how in stroke or out of stroke they each were, etc, etc, etc, etc, etc, the formula becomes too complicated, and there would be way too much data to input. Mike would need a team of 50 data entry people to keep up with it even if the formula could accurately take all these things into consideration.

I’m also pretty certain it is hard enough to get league operators and tournament directors to report all their match statistics as it is right now when all they have to report for each match is the identity of the players, who won, and the scores. Many people are inherently lazy, especially if they don’t feel that something is directly benefitting them right now at this moment. Can you imagine if they had to also report all table conditions, or the weather conditions, or the conditions of the opponents (drunk, in stroke, getting over the flu) too? Forget about it, nobody would ever submit any data to Mike if it got much more involved than the players and the score. Not to mention that much of that other stuff is totally subjective anyway and whatever they reported wouldn’t even be reliable. These are just a few of the many examples of how “improving” any one thing is going to make something else much worse.

The fact of the matter is that FargoRate is about as perfect as is possible for something which is intended to be able to rate and compare all the pool players of the world against each other. Any little thing you could make a little better in one way just makes something else much worse. Right now it is about the perfect all around compromise as it is, and the nice thing, should Mike ever see the need for it and want to have a second or third “version”, is that FargoRate could be tweaked to be even a little bit better for any very specific application, such as for 8 ball only, or if used just for the pros.

The one (and really only) thing that can make it better right now as a system for every pool player in the world, and which everyone already universally agrees on, is more data. The more match statistics there are for everyone, the better and more accurate it gets (and it is already pretty darn good). So the best thing, and really the only thing that can be done to improve it at this point is for all of us to ask every league and tournament director to submit all their match statistics to Mike for inclusion.

I think because the pros have so much data available they are good to use as the benchmark from which all else flows.

SVB for example has 8000+ games in the system. Archer has 5000+

Most amateurs aren't going to have that much in. But when the forumula is applied to those with the most data and the predictions seem to be working out as they should then we can be reasonably certain that the formula works for everyone else as well.

And as with everything...over time improvements become clear.
 
^^^^^^^^^^^

That was a very thorough and well written comment above. I enjoyed the input.

If I had my wish list, #1 would be to factor Dr. Dave's TDF into the match as it's entered into the Fargo Rate system. I still believe 7-foot table results should be weighted differently than the big tables-or more specific matchup tables (like the ones at hard times and Steinway with less than 4" pockets).

I don't play golf but I understand there is a central computer system that an amateur enters his results in to prove his handicap. I also understand the particular course this amateur just played has a difficulty rating that is factored in. So it makes sense to use the tables difficulty rating in pool.

Golf courses have a course rating and a slope rating, allowing you to estimate how much more difficult a particular course is to golfers of different skill. This is needed because your golf score is an absolute score relative to a particular course on a given day. What that simple scheme does not consider is that some days are windy, other days the grass is or is not freshly mowed, the ball carries very different distances when the ground is dry and hard than when it is soft.

If Fargo Ratings were applied to golf, we would never, ever record a golf score. You would play against an opponent of known rating, and we would record, for each hole, how many strokes you took relative to your opponent. So today it is raining and you are playing Tiger Woods at Winged Foot, and you scored +1, +3, 0, +2 on the first four holes. Tomorrow you play your brother-in-law on a different course and score 0, -1, 0, -1. Note that it will take a long time (a lot of data) to get things sorted out. But when things ARE sorted out, the ratings will have considered not only course difficulty and slope ratings, but also all those other things that might vary from day to day and are not in the simple description of course difficulty.

To be sure there are subtle differences in how different golfers are affected by different conditions. For example, very windy is worse for a golfer who relies on hitting the ball far. Wet, soft ground is worse for the golfer who doesn't hit the ball far (and relies on more distance after the drive hits the ground). Neither approach accounts for these subtle differences. And that is where we are at with Fargo Ratings for pool. We do not take into account subtle differences that might benefit one player over another on different equipment.

But that doesn't mean Fargo Ratings don't take table-difficulty into account. They actually do in a way that that considers more than a simple table difficulty factor. Your score relative to your opponent's score takes into account the size of the table, the size of the pockets, the cut of the pockets, the speed of the cloth, whether the rail cloth is slick making the pockets play a little bigger, whether there are rolls on the table, whether one rail is deader than the others..... This, in effect, is all in there for Fargo Ratings.
 
Golf courses have a course rating and a slope rating, allowing you to estimate how much more difficult a particular course is to golfers of different skill. This is needed because your golf score is an absolute score relative to a particular course on a given day. What that simple scheme does not consider is that some days are windy, other days the grass is or is not freshly mowed, the ball carries very different distances when the ground is dry and hard than when it is soft.

If Fargo Ratings were applied to golf, we would never, ever record a golf score. You would play against an opponent of known rating, and we would record, for each hole, how many strokes you took relative to your opponent. So today it is raining and you are playing Tiger Woods at Winged Foot, and you scored +1, +3, 0, +2 on the first four holes. Tomorrow you play your brother-in-law on a different course and score 0, -1, 0, -1. Note that it will take a long time (a lot of data) to get things sorted out. But when things ARE sorted out, the ratings will have considered not only course difficulty and slope ratings, but also all those other things that might vary from day to day and are not in the simple description of course difficulty.

To be sure there are subtle differences in how different golfers are affected by different conditions. For example, very windy is worse for a golfer who relies on hitting the ball far. Wet, soft ground is worse for the golfer who doesn't hit the ball far (and relies on more distance after the drive hits the ground). Neither approach accounts for these subtle differences. And that is where we are at with Fargo Ratings for pool. We do not take into account subtle differences that might benefit one player over another on different equipment.

But that doesn't mean Fargo Ratings don't take table-difficulty into account. They actually do in a way that that considers more than a simple table difficulty factor. Your score relative to your opponent's score takes into account the size of the table, the size of the pockets, the cut of the pockets, the speed of the cloth, whether the rail cloth is slick making the pockets play a little bigger, whether there are rolls on the table, whether one rail is deader than the others..... This, in effect, is all in there for Fargo Ratings.

This is a great answer and I thank you for the extra information.
I appreciate what you're doing for the game. Labor of loves are never easy. My hat is off to you.
 
Back
Top