Fargorate VS Pro Player designed rating system

And either way, you’re still trying to fix a cart with no horse. Nothing in any of your posts suggests ways to put people in bleachers or behind a tv.

Until someone fixes that, you’re wasting time with anything else.
 
Also, as others have stated, just because it’s not specifically taken into account, many things are covered by Fargo without needing to be specifically addressed.

If someone(any age) is losing because they don’t have good stamina, their rating will go down because they are losing, regardless of the reason.

So, that point is moot.

Also, adding more time into tournaments will keep them from being able to be played on a weekend(except the major ones like the Open that already take a week and don’t have people playing as long). That takes away some of the dead money recreational players and also pro level players who have a full time job. Your suggestions hurt pool in its current state more than it helps.

Other issues are far more important.
 
strmanglr scott said:
Do the pockets know the age, sex, race, religion etc.of the person putting balls in?
Click to expand...
No, but father times does and he can be mean sometimes. Ask me how I know.
On another note, I'm sure FargoRate will see the decline in a player ability and adjust to the players as more info goes in the system. jmo
 
Pacecar said:
As Mike Page explained before, FargoRate has an age declining function. Older games in the database don't have as much weight as current games.

One-year-old games are worth about 80% of a game played today

1 year: 80%
3 years 50%
6 years 25%
10 years 10%

http://forums.azbilliards.com/showthread.php?t=438483&page=14
See posting #210
Click to expand...

Ya so a player player has a rating of 780 3 yrs ago and he comes back plays in a few events goes 2 and out losing a average of 9-5 to the same level players where then is his current rate going to slide to

1
 
one stroke said:
Although Fargo might be a probably is the most accurate you still could have some problems when using it for rail betting , for instance James Aranas Zoren hit the scales as number 3 at 813 I believe most of it based on him playing this person or that person mostly not here , when he came here and I saw him play I told his handler after he made mention of that he's not that speed , he's now 802 ,he had a low robust that would explain that
However take Alex with a high robust number who is still down the list because Alex does what Alex does plays many games one pocket ,snooker cards what ever might jump into some rotation also so his number stays lower , then he might just decide to dedicate all his time strictly to Rotation and in a short time he's playing 810 speed but his number won't reflect that similar to a batting average, you see that in the summer after a guy hits 225 then he's hitting 360 and his number rounds out to 280 and he's simply a better hitter than that ,
Without being privy to that info you simply don't have a today accurate rate period

1
Click to expand...
All you essentially said above is that short term factors such as who has or has not been playing a lot in the last month, or who has the flu, or who doesn't have enough games in the system for us to reliably know what their true average speed is yet, or who is going through a divorce, and other such short term factors can have an effect on how someone is going to play on any given day and therefore should be taken into consideration if you are trying to bet on a match. I think everybody absolutely agrees with you and wonders what your point is since this is so obvious.

It is no different in anything else either. Take football for example. Which team is more highly rated means a lot when it comes to knowing who is likely to win the game, but which players are out on the injured list can have a big impact on determining who wins too. A rating system not knowing about or being able to predict those types of short term factors is not in any way a deficiency with the rating system.
 
Pacecar said:
As Mike Page explained before, FargoRate has an age declining function. Older games in the database don't have as much weight as current games.

One-year-old games are worth about 80% of a game played today

1 year: 80%
3 years 50%
6 years 25%
10 years 10%

http://forums.azbilliards.com/showthread.php?t=438483&page=14
See posting #210
Click to expand...
For comparison the world golf rankings drop results completely after 2 years.

I know 10% doesn't contribute much to the ranking, but this still seems too long given how much things can change in 10 years. Mike, have you done any testing on optimal
dropout rates for old match results?
 
skip100 said:
For comparison the world golf rankings drop results completely after 2 years.

I know 10% doesn't contribute much to the ranking, but this still seems too long given how much things can change in 10 years. Mike, have you done any testing on optimal
dropout rates for old match results?
Click to expand...

One problem with dropping results after two years is then you don't know anything about people that stop competing for extended periods of time and then come back.
This happens a lot in sports, but probably even much more so in pool than in many others. So when you quit playing for say 6 six years and then come back, yes you aren't going to be as good as you were six years ago at first, but you aren't going to suck either, plus your game is likely to improve very rapidly. And since FargoRate so heavily weights your most recent results, it quickly "overrules" your old data and very rapidly adjusts you rating to be very close to your true current abilities. There is a strong argument to be made that having rating info on someone, even if it is slightly off right at first for people that haven't played for an extended period of time, is loads better than having no info at all.

Another benefit to counting results that go back so many years is that it allows you to get a larger sample size of people's play. If everyone was competing once a week, and if all tournaments and leagues reported their results to FargoRate, then you would probably have enough info for super accurate ratings even if you only went back a year, but neither of those things are the reality. Most tournaments and leagues still aren't reporting their results to FargoRate and so FargoRate is having to rely on the results from just a small percentage of them. Probably an even bigger factor is that many players just don't compete regularly. Many just compete a time or two a year on average if that, and so you just don't have enough info for accurate assessments on them unless you collect information over a longer period of time. And there are many others that tend to compete in spurts. They may compete fairly regularly for a little while, and then they take one, two, three, or more years off from competing or playing, and so again you often just wouldn't have much info on these people much of the time if you only went back a couple of years.

All things considered, it seems like it makes the most sense to count results that go back quite a number of years, but weigh the more recent results much more heavily, thereby allowing you to get the best compromise between how current and accurate your ratings are, and the number of people you can provide pretty accurate ratings for. Essentially you could have super accurate ratings for a very small portion of players, or really accurate ratings for most of the players, and which of the two is best seems pretty clear especially when you consider that being able to accurately rate and be used by as many players as possible around the world was one of the main goals of FargoRate to begin with--one system that everybody can use regardless of location.
 
fargo rate

I think fargorate is very good, and a work in progress.
most poolplayers are excellent complainers...nuff said LOL
 
You must log in or register to reply here.
Back
Top