Updated FargoRates are out

There is a big difference in a tournament match and gambling match race to 30, 50, 80, or even 100. BIG difference... especially for pool. At 740 and above race to 9 or even 11 a 740 has a really really good chance at beating anyone in the world but in reality they aren’t even in the top 200 players in the world. That’s why races to 7,9 etc. are really a joke and that’s what we play in our sport and that’s why we have 20 different winners out of 20 tournaments. You would never ever see a 740 beat Orcollo in a real pool match. A race to 9 in any game other than one pocket is not a real pool match IMO.

I think Fargo is better than any other handicap system but sometimes it just seems so far off that I couldn’t use it. I would love to bet on a lot of the matches that are predicted. Our game is way to inconsistent. We need to be using the same cue ball every weekend for one... For a month a player uses a measle, then flys to Vegas and now he had to use cyclop cue ball... I guarantee It will affect the better player. He may adjust faster but the better player plays a more precise game and if he is dialed and has been training all month with a different cue ball that’s huge.... tables speed, rails, etc. is one thing but cue ball never gets brought up.. All that stuff matters... I honestly think some guys have a good chance winning with certain racks, certain cue balls, etc. and their chances can go down big time if they aren’t using equipment they have been training with.

I think Donny would win easily by 20-30 games in a race 100 against Siming Chen playing 9 or 10 ball race to 100 if both players got a month to practice with whatever they were going to play on.
 
Justin Bergman said:
I would love to bet on a lot of the matches that are predicted.
Click to expand...

Once more, Fargo Rate isn't meant to predict matches.

You can use it to set lines if you want, but not predict the outcome.

If a player rated 800 plays someone rated 740, the 800 is a 3 game favorite in a race to 11.

This simply means the line would be -3.

It doesn't mean the higher rated player would always cover the spread, or that they would never lose. After all, the undefeated Patriots did lose the Superbowl to a Giants team that almost missed the playoffs.
 
Justin Bergman said:
I think Fargo is better than any other handicap system but sometimes it just seems so far off that I couldn’t use it.

I think Donny would win easily by 20-30 games in a race 100 against Siming Chen playing 9 or 10 ball race to 100 if both players got a month to practice with whatever they were going to play on.
Click to expand...

this and this!

i also think that fargo is in some spots just unreal and that makes the whole system invalid imo latest joke was dodong, played 2 or 3 tourneys and gets no3 spot (rhea, how much money do you have won last 2-3 tourney betting on dodong :grin-square: )

and to think that siming chen would stand a chance vs the likes of donnie under any circumstances is a bit delusional! she might pull off one or another rt 9 against them, but she never overcomes them in the long run or in a long gambling match!
 
Last edited:
I like the fact that Donny didn't focus on Aiming Chen (<- my autocorrect changed Siming to Aiming, but I like it), and instead brought up other male players rated similarly.

The fact that disagreement with her rating is wrapped up with criticism of Fargo ratings in general makes it hard to separate a bunch of guys underestimating a girl, from a serious criticism of a mathematical rating system that's over the heads of most of us. Sorry, but I'm going with the math. Even if I don't really understand it, I've seen Mike Page's posts about it enough to be convinced that it's sound.
 
Justin Bergman said:
What’s our game Donnie Mills? You break better than me in every game but I will play any game you want. Last time we played I got the last 5 lol.
Click to expand...

Lol that was 14 years ago. Now you're gonna have to spot me. And I'll take games on the wire. Probly 8 going to 50 and I don't think I like it.
 
Justin Bergman said:
There is a big difference in a tournament match and gambling match race to 30, 50, 80, or even 100. BIG difference... especially for pool. At 740 and above race to 9 or even 11 a 740 has a really really good chance at beating anyone in the world but in reality they aren’t even in the top 200 players in the world. That’s why races to 7,9 etc. are really a joke and that’s what we play in our sport and that’s why we have 20 different winners out of 20 tournaments. You would never ever see a 740 beat Orcollo in a real pool match. A race to 9 in any game other than one pocket is not a real pool match IMO.

I think Fargo is better than any other handicap system but sometimes it just seems so far off that I couldn’t use it. I would love to bet on a lot of the matches that are predicted. Our game is way to inconsistent. We need to be using the same cue ball every weekend for one... For a month a player uses a measle, then flys to Vegas and now he had to use cyclop cue ball... I guarantee It will affect the better player. He may adjust faster but the better player plays a more precise game and if he is dialed and has been training all month with a different cue ball that’s huge.... tables speed, rails, etc. is one thing but cue ball never gets brought up.. All that stuff matters... I honestly think some guys have a good chance winning with certain racks, certain cue balls, etc. and their chances can go down big time if they aren’t using equipment they have been training with.

I think Donny would win easily by 20-30 games in a race 100 against Siming Chen playing 9 or 10 ball race to 100 if both players got a month to practice with whatever they were going to play on.
Click to expand...



This is exactly right!!! This is music to my ears! Finally I read something in this thread that makes sense! For all of you people on this forum that aren't sure what's accurate or not. This is accurate!
 
BeiberLvr said:
Once more, Fargo Rate isn't meant to predict matches.

You can use it to set lines if you want, but not predict the outcome.

If a player rated 800 plays someone rated 740, the 800 is a 3 game favorite in a race to 11.

This simply means the line would be -3.

It doesn't mean the higher rated player would always cover the spread, or that they would never lose. After all, the undefeated Patriots did lose the Superbowl to a Giants team that almost missed the playoffs.
Click to expand...

According to Fargo the longer the race the more chance the better player wins. Which makes perfect sense. Amar Kang is a 766 and I'm a 753. Fargo says I have a 28% chance to win in a race to 80. I'll give him the last 2 and make my odds even worse. If Fargo has us rated so accurately why doesn't somebody want this free money?? Do you know how hard it is to get your money in a 80% favored situation??
 
Justin Bergman said:
That’s why races to 7,9 etc. are really a joke and that’s what we play in our sport and that’s why we have 20 different winners out of 20 tournaments.
Click to expand...

Well, that is simply not true. I don't think, that there are more than 10 different players, that have won one of the last 20 EuroTour events.
Same goes for the DCC 9ball or the US-Open....ok, US-Open might have had more winners, but even over a 20 YEAR long timespan there have been multiple wins by SvB, Darren and Mika.
Only the WC nearly had 20 different winners in 20 years(i think Hohmann was the only one who won 2 titles).

So either the short race hasn't that much of an impact, or people like Shane or Ralf Souquet, who has had something about 45 top three ET results, are not good but lucky.

So you could argue, that the player who won the race to 9 match won't win the race to 50, but that is kinda like telling Usain Bolt that he might be the fastest sprinter on earth, but to proof it he has to run a 5 km race.



I totally agree on your point, that pool is too inconsistent. It is not only about balls, taples, cloth etc.
Pool is missing an international organisation, that regulates everything. Starting with official equipment, over official tournaments to official pro status.
Today we have pool in Europe, in the U.S, in Asia and one or two times a year the top players come together for a tournament...or at least some of them.
Fargo ratings are a result of that inconsistency, because we want to compare the Taiwanese superstar, with the US gambler and the European tournament player...with the chance, that this three players won't even play against each other in their whole career.
 
BRussell said:
I like the fact that Donny didn't focus on Aiming Chen (<- my autocorrect changed Siming to Aiming, but I like it), and instead brought up other male players rated similarly.

The fact that disagreement with her rating is wrapped up with criticism of Fargo ratings in general makes it hard to separate a bunch of guys underestimating a girl, from a serious criticism of a mathematical rating system that's over the heads of most of us. Sorry, but I'm going with the math. Even if I don't really understand it, I've seen Mike Page's posts about it enough to be convinced that it's sound.
Click to expand...

maybe i can help you to understand!
lets take busti, hes a good excemple, he was top 15 til a few weeks ago
do you know that busti (efren aswell) got 2 games to 22 from 18 yr old, unranked anton raga already about 2 yrs ago? the same kid who got spots from the top philipnos at the same time? that makes busti a top 15......in the philipines!
i just trust those matchmakers, who do nothing else all day, much more than a system who pulls racks together without regarding the circumstances!

.....another great excemple is MD

Mike Dechaine USA795
Eklenti Kaci ALB795
Joshua Filler GER795
Ping Chung Ko TPE794

who would you bet on matching MD up even with the 3 guys behind him? that match ups are a betters paradise when you find someone backing MD lol

or efren


Ronnie Alcano PHI794
Justin Bergman USA793
Lee Vann Corteza PHI793
Albin Ouschan AUT792
Mark Gray GBR792
Efren Reyes PHI791
Yu Lung Chang TPE790
Kun Lin Wu TPE790
Darren Appleton GBR790
Yu Hsuan Cheng TPE790

make those matches happen and i bet houses (ok, i would if i had some :grin-square: ) against efren every single match!

or mosconi! fargo has them pretty even and the bookies 3/1 (might have dropped a bit, havent checked so far) , even with homefield advantage.....again, who to trust? the guys who hate losing money or the system which just pulls racks together?

or albin? winning everything last 3 years and is slowly crawling to the top 20 while dodong plays 3 tourneys to get no3 spot!....... is albin upsetting the poolworld again and again with his "lesser skillset" or is he simply underrated?

sry, but i cant take that serious and i would NEVER handicap players according to it, way too many bad spots in it (and ONE bad spot already makes it unfair, and there are alot, just like siming)!
 
donny mills said:
According to Fargo the longer the race the more chance the better player wins. Which makes perfect sense.
Click to expand...

yes.

Amar Kang is a 766 and I'm a 753. Fargo says I have a 28% chance to win in a race to 80.
Click to expand...

No. You've said this a few times, so I think it is time to explain. FargoRate does not think you have a 28% chance of beating Amar in a race to 80. There is an important subtlety here, one that can usually be ignored but cannot here. I'll get to that subtlety, but it will take a bit.

So first, what does the % calculator say? What is it actually telling you? Forget assigning ratings to players for a minute. Imagine we have a hypothetical player who plays at exactly 766 speed and another hypothetical player who plays at exactly 753 speed. We don't have that, but imagine we did. A single game between these players has a 52.3% chance of being won by the 766 and a 47.7% chance of being won by the 753.

Here is an analogy. Playing a single game is like fishing an M&M at random out of a bag that contains 1000 M&Ms, 523 of which are red and 477 of which are brown. So you fish out an M&M, record it's color, and throw it back in. If we do a race to 80, and I take red, I will win 72% of the time. That's what the calculator says, and the calculator is right about that.

Note, here, that we know exactly how many red and brown M&Ms are in the bag. How do things change if we have some uncertainty about the how many of the 1000 M&Ms are red. Perhaps it is not 523. Perhaps it is 528. Or perhaps it is 518. The usual situation is the likelyhoods we are off in one direction (overestimate reds) and likelyhood we are wrong in the other direction (underestimate reds) are the same, and one bumps up red's chances and the other bumps down red's chances and we get a canceling effect. If we rely on this canceling, maybe our original conclusion about the 72% remains the same even in the face of the uncertainty in the M&M distribution.

You've chosen a situation--a long race-- in which this canceling of the effects of uncertainty doesn't work. And that makes the 28% estimate wrong. I'll make an extreme example to illustrate. Imagine we have two players rated 700 and 720, and we are certain about their ratings: there really is a 20-point gap. If they play a super long race, like a race to 10,000, the 700 has essentially zero chance of winning.

Now suppose there is some slop in their ratings such that our assessment of the rating gap looks like this: instead of a 100% chance the rating gap is 20 points, there is now a 50% chance it is 20 points and a 25% chance each it is 0 or 40 (20 points lower or 20 points higher). Now when we look at the chance the lower-rated player wins a race to 10,000 we have to take all three scenarios into account. If the rating gap is 20, then like before, the 700 has virtually no chance. If the rating gap is really 40, then the chance of the 700 winning went from really small to really really small. But if the rating gap is 0, then the chance of the 700 winning becomes 50%. So in the overall picture, there is a 25% chance of the match being 50/50, which means there is a 12.5% chance of the lower-rated player winning a race to 10,000 when there is a 20-point gap in the ratings and we introduce uncertainty in the ratings like described here.

Forget Donny and Amar for a minute, the race-to-80 calculator for the above reason exaggerates the advantage of the higher-rated player for ANY two players with a 13-point gap. Essentially in a long race, the higher-rated player has nothing to gain and everything to lose from uncertainty in the ratings.

If I am your financial advisor, I am suggesting you bet on the lower-rated player every time in a long race if you are given odds that comes from those percentages in the calculator.

I'll give him the last 2 and make my odds even worse. If Fargo has us rated so accurately why doesn't somebody want this free money?? Do you know how hard it is to get your money in a 80% favored situation??
Click to expand...

What do you mean by "if Fargo has us rated so accurately"? We've actually been pretty silent on the nature and size of the uncertainty in the ratings. That is a complicated issue, and it depends on a lot of things--how many games you've played, how recent they have been, how well rated are each of your individual opponents, how many of your games are against similarly rated opponents vs opponents much better or much worse, how many of your games are played in run-out conditions where many games have one inning.
We've actually said very little on the subject outside of whatever the uncertainty is, more data reduces it.
 
pmac666 said:
maybe i can help you to understand!
lets take busti, hes a good excemple, he was top 15 til a few weeks ago
do you know that busti (efren aswell) got 2 games to 22 from 18 yr old, unranked anton raga already about 2 yrs ago? the same kid who got spots from the top philipnos at the same time? that makes busti a top 15......in the philipines!
i just trust those matchmakers, who do nothing else all day, much more than a system who pulls racks together without regarding the circumstances!

.....another great excemple is MD

Mike Dechaine USA795
Eklenti Kaci ALB795
Joshua Filler GER795
Ping Chung Ko TPE794

who would you bet on matching MD up even with the 3 guys behind him? that match ups are a betters paradise when you find someone backing MD lol

or efren


Ronnie Alcano PHI794
Justin Bergman USA793
Lee Vann Corteza PHI793
Albin Ouschan AUT792
Mark Gray GBR792
Efren Reyes PHI791
Yu Lung Chang TPE790
Kun Lin Wu TPE790
Darren Appleton GBR790
Yu Hsuan Cheng TPE790

make those matches happen and i bet houses (ok, i would if i had some :grin-square: ) against efren every single match!

or mosconi! fargo has them pretty even and the bookies 3/1 (might have dropped a bit, havent checked so far) , even with homefield advantage.....again, who to trust? the guys who hate losing money or the system which just pulls racks together?

or albin? winning everything last 3 years and is slowly crawling to the top 20 while dodong plays 3 tourneys to get no3 spot!....... is albin upsetting the poolworld again and again with his "lesser skillset" or is he simply underrated?

sry, but i cant take that serious and i would NEVER handicap players according to it, way too many bad spots in it (and ONE bad spot already makes it unfair, and there are alot, just like siming)!
Click to expand...

Time is certainly a confounding factor that the algorithm might not properly address. Fargo or ELO is a lagging indicator. Thus, unless events actually happen, prediction (or indication of prediction) does not change or it might only change slightly when global optimization is performed.

I've seen the implementation of Fargo in USAPL and thought it was genuinely accurate. In my humble and professional opinion, the effect might due to players in the same league are often very heavily connected, thus making predicting match outcomes "easier". For the example between Donny and Chen, I'd be curious to see if there's any difference if Mike only uses cohorts that both players have played in the past.
 
Last edited:
donuteric said:
Time is certainly a confounding factor that the algorithm might not properly address. Fargo or ELO is a lagging indicator. Thus, unless events actually happen, prediction (or indication of prediction) does not change or it might only change slightly when global optimization is performed.

I've seen the implementation of Fargo in USAPL and thought it was genuinely accurate. In my humble and professional opinion, the affect might due to players in the same league are often very heavily connected, thus making predicting match outcomes "easier". For the example between Donny and Chen, I'd be curious to see if there's any difference if Mike only uses cohorts that both players have played in the past.
Click to expand...

yeah, i can see that it might work better in regional leagues where players are connected and in similar pressure situations......
but worldwide? with bartables included? they always say it evens out with time but shouldnt they have enough data now, esp for players like busti, efren or MD, to show the current speed and not the speed 3 yrs ago or longer? efren will still have a 780 rating 5 years after he passed lol
also IT IS a HUGE difference winning world 9b or a 2k added bartable event!
nope, i dont buy it (but who am i? lol)....i anyway think that handicapping in tourneys is nonsense!
 
mikepage said:
yes.



No. You've said this a few times, so I think it is time to explain. FargoRate does not think you have a 28% chance of beating Amar in a race to 80. There is an important subtlety here, one that can usually be ignored but cannot here. I'll get to that subtlety, but it will take a bit.

So first, what does the % calculator say? What is it actually telling you? Forget assigning ratings to players for a minute. Imagine we have a hypothetical player who plays at exactly 766 speed and another hypothetical player who plays at exactly 753 speed. We don't have that, but imagine we did. A single game between these players has a 52.3% chance of being won by the 766 and a 47.7% chance of being won by the 753.

Here is an analogy. Playing a single game is like fishing an M&M at random out of a bag that contains 1000 M&Ms, 523 of which are red and 477 of which are brown. So you fish out an M&M, record it's color, and throw it back in. If we do a race to 80, and I take red, I will win 72% of the time. That's what the calculator says, and the calculator is right about that.

Note, here, that we know exactly how many red and brown M&Ms are in the bag. How do things change if we have some uncertainty about the how many of the 1000 M&Ms are red. Perhaps it is not 523. Perhaps it is 528. Or perhaps it is 518. The usual situation is the likelyhoods we are off in one direction (overestimate reds) and likelyhood we are wrong in the other direction (underestimate reds) are the same, and one bumps up red's chances and the other bumps down red's chances and we get a canceling effect. If we rely on this canceling, maybe our original conclusion about the 72% remains the same even in the face of the uncertainty in the M&M distribution.

You've chosen a situation--a long race-- in which this canceling of the effects of uncertainty doesn't work. And that makes the 28% estimate wrong. I'll make an extreme example to illustrate. Imagine we have two players rated 700 and 720, and we are certain about their ratings: there really is a 20-point gap. If they play a super long race, like a race to 10,000, the 700 has essentially zero chance of winning.

Now suppose there is some slop in their ratings such that our assessment of the rating gap looks like this: instead of a 100% chance the rating gap is 20 points, there is now a 50% chance it is 20 points and a 25% chance each it is 0 or 40 (20 points lower or 20 points higher). Now when we look at the chance the lower-rated player wins a race to 10,000 we have to take all three scenarios into account. If the rating gap is 20, then like before, the 700 has virtually no chance. If the rating gap is really 40, then the chance of the 700 winning went from really small to really really small. But if the rating gap is 0, then the chance of the 700 winning becomes 50%. So in the overall picture, there is a 25% chance of the match being 50/50, which means there is a 12.5% chance of the lower-rated player winning a race to 10,000 when there is a 20-point gap in the ratings and we introduce uncertainty in the ratings like described here.

Forget Donny and Amar for a minute, the race-to-80 calculator for the above reason exaggerates the advantage of the higher-rated player for ANY two players with a 13-point gap. Essentially in a long race, the higher-rated player has nothing to gain and everything to lose from uncertainty in the ratings.

If I am your financial advisor, I am suggesting you bet on the lower-rated player every time in a long race if you are given odds that comes from those percentages in the calculator.



What do you mean by "if Fargo has us rated so accurately"? We've actually been pretty silent on the nature and size of the uncertainty in the ratings. That is a complicated issue, and it depends on a lot of things--how many games you've played, how recent they have been, how well rated are each of your individual opponents, how many of your games are against similarly rated opponents vs opponents much better or much worse, how many of your games are played in run-out conditions where many games have one inning.
We've actually said very little on the subject outside of whatever the uncertainty is, more data reduces it.
Click to expand...



Yeah but like what if it rained the day before?
 
pmac666 said:
maybe i can help you to understand!
l[...]
.....another great excemple is MD

Mike Dechaine USA795
Eklenti Kaci ALB795
Joshua Filler GER795
Ping Chung Ko TPE794

who would you bet on matching MD up even with the 3 guys behind him? that match ups are a betters paradise when you find someone backing MD lol
Click to expand...

Two of those players are rated above Dechaine, not below him.

Regardless, they're all in the same ballpark so it doesn't much matter who is a few points above who.

Why do you think Dechaine would underperform? Is it because he hasn't been in the ring much?

Three weeks ago he played Yu Lung Chang (792) 44 games. Mike won 23 and lost 21 --about what we'd expect. And in that tournament generally he performed at 810 speed for 109 games.

I'm wondering what is the basis for your thinking about Mike.

And Dodong now has over 600 games performing at 802 speed.

It's amazing. As soon as one player spots another player a little something in a set that could go either way even with the spot in the other direction, that becomes some sort of fundamental knowledge that the spotter plays better than the spottee... You guys should be a little more careful with your conclusions and just a little more cynical about the legendary army of world-class 12-year-olds in the Philippines. I'm not saying there are not several very straight-shooting young players in the the Philippines; we actually have evidence there are. But ask yourself, if these stories that are repeated over and over again get embellished just a little bit here and there which direction are they going to morph? --like the number of pool rooms in Shanghai. I think somebody is confusing pool rooms and Starbucks.
 
pmac666 said:
shouldnt they have enough data now
Click to expand...

No, they don't. Far from it. From the website's statistics, 6 million games from 127k players means the average is far less than how the algorithm would define robustness as established. Assuming the 6 million games are total of everything, they probably have much less if they only include "recent" games.

Data collection is often far more challenging than algorithmic development and it requires partnerships from promoters, directors, and operators. My understanding is Mike et al are doing the best they can, but from a data scientist's point of view, 6 millions is a relatively low number in the big data era.
 
Last edited:
mikepage said:
We would be very interested to see a long race between you and Siming Chen.

As far as us betting on the outcome, please understand that because we control the data input and the integrity of the data generally, we as a matter of policy don't bet on pool matches.

And to a few of the others who see some potential match as being some sort of test of Fargo Ratings, that's just silly. There are already millions of games of "tests" of the FargoRate system. The idea that one or two races to 50 or something somehow provide some sort of test is ridiculous.
Click to expand...

You really don't believe she plays close to donnys level?
 
Seems like the system is too granular. What's the point of calling someone a 763 when they could be plus or minus 20 points? Maybe it needs a +/- for 1 or 2 standard deviations out next to the rating.

Then what's the use of the % chance to win race calculator, if it's highly dependent on the length of race and the possible variation in someone's rating?
 
real bartram said:
You really don't believe she plays close to donnys level?
Click to expand...

It depends on what we mean by close. She may. If she's a little overrated and Donny's a little underrated, they of course would move toward one another.

But I think Siming plays better than Donny.

And I think the vast knowledge and experience of a handful of you who are smart and make good games and have a finely tuned intuitive sense of how people play relative to one another is a little out of its element when it comes to the top Asian women.

I see Donny as believing he is a little underrated. And I see him perusing the list for someone above him who he thinks is a little overrated and proposing a game with a modest spot. That's good and everything. I always like to see more games go down. But it is not earth shattering.

Here is I think a more interesting experiment. Take the USA top 50 list and generate a couple dozen random pairings: Oscar vs Sossei, Sky vs Josh Roberts, and on and on. And then give that list to a handful of people like you who others consider knowledgable. Ask two questions... Who will win a race to 30? How many games will the loser win? I think averaging over a number of knowledgable people, the results would be close to the expectation from FargoRate.
 
You must log in or register to reply here.
Back
Top