I guess that if your Fargo robustness is low, you probably don’t want to join a league where the vast majority of players in it are lower rated than you when you start off. I’ve played pool for 40+ years & I have a good idea of my skill level and it sure isn’t 660+. This rating will only hurt if I enter a team tourney & that’s if I can find a team where that rating doesn’t cause difficulties. Oh well.
Can I question my Fargo rating?
- Thread starter roger.p
- Start date
Yeah, your math is off. A 75% win rate means if you play 4 games, you win 3, or 3:1. 100 points is 2:1. To get 3:1, you need 150 point difference. The calculator shows 160 points is slightly better than 3:1. I think you're confusing 75 points with 75%.
I was going by a 100 point difference in double the wins needed, 75 of 100 is 75%, but it's not quite how the games work out.
I think all this proves is the OP isn't lying with the data provided. OP's concern I think is that doesn't believe he's really a 50/50 shot vs. the other Fargo 660's in the world.
I can see a situation where if you either are the best in your population or the worst in your population by skill set, that a numerical rating system may cause rating inflation or deflation, as it's trying to extrapolate from a core data set that is distant from the OP's own skillset. It looks like in the OP's case the upper bound of his skill isn't being tested in his current population.
Another alternative (maybe): The increasing use of FargoRate in leagues (e.g. BCAPL, USAPL) is pumping a lot of lower rated players into the system, which has the statistical impact of moving the right hand side of the performance distribution curve further right. There are 33 Fargo 800+ players now vs. the "around 20" in the description of Fargo. For instance SVB now is an 834, however I don't think people will make an arguement that he is better skill-wise now than when he was ruling the world as a Fargo 818 in 2016.
It's not so much having lower rated players, it's the fact you don't have a lot of games in and you are only playing a single game against them. You need a lot of single games to end up with a correct % of wins over time to even out.
FargoRate is a rating system, not a ranking system. The two kinds of "player performance" systems are very different.
A ranking system only looks at recent competition and awards points related to performance. Snooker has a very simple system: How many British pounds have you won in the last two years in ranking events? The UMB awards ranking points based on finishes in tournaments and I believe they also use a two-year cycle, but they do not use prize money directly.
A rating system is different. It tries to give you a good estimate of a player's ability. Elo ratings are one example of this. They try to estimate the chance that you will beat your opponent based solely on previous games won/lost and the ratings of your opponents.
It is reasonable for a rating system to count recent performance more heavily than games played years ago and FargoRate does that. It may be on the FR FAQ page. However, if someone has not played in recorded matches for five years you have no information about whether they play better or worse. FR leaves their rating alone in the absence of knowledge. If they start playing again, their rating will adjust fairly quickly due to the "aging" of old data. For the same player, most ranking systems would rank them as "cannot play at all" due to their absence from ranking event play.
You can sort out your "correct" rating by judging your skill at the table if the players are not enough for you to comete well against also. I have seen a lot of players and what they can do, my son is a 630, I have seen him run 3-4 racks in a match in league and quite often he gets a few break and runs in tournaments. I'm a 550, I have done two racks in a row several times and can do a few run outs in a night of playing on decently tough equipment. My average ghost play in 9 ball is 7 to 7.5 balls, I can beat the 7 ball ghost but not the 8 ball ghost.
A 650 player or so is a solidly A player that can run out many open racks.
If he has enough games (and perhaps he doesn't), that works itself out fine. Justin Sajich from Australia (765 and almost world top 100) first came to our attention when he had only league games from Diamond League in Sydney and he was at 760.
I think all those top dogs are playing better now. If we look at the top 1000 players in FargoRate from 4 years ago that had 800 or more games in the system and compute an average rating we get 698.3. If we look at that SAME 1000 players now, they have an average rating of 699.8. They've stayed the same. No rising tide.
How do they rate their beginners and not so good players? According to their ratings, they need an F rating for a 400 Fargo. A bit messed up ratings there since I see D and C players getting to the quarter finals and semifinals against A and B players playing even. There is a C beating a B 7-4 and then going hill-hill with an A. Pretty impossible if the ratings are correct for the skill levels.
Last edited:
They're not playing even. The race is to 7, and there is a chart for the ratings that says a lower-rated player can start, for example, at 2 instead of at 0. That's the way cuescore does it.
They COULD go to F. They also have a few other posibilities in Poland ;-)
This is eye opening. A knowledgeable 620 bar table 8 ball player should have no problem winning at a 80% rate against 500 fargo rated players imo. Guess my opinion is way off lol
Your robustness is 221. That is a very low number. Which means that before you we established your number was essentially a guess. most start at 525 just as a placeholder and their rating will fluctuate from there. But it could start at 575 depending on what information they have or a TD can start you out at whatever rating they think fits. Once you hit 200 the rating gets established. So if you were a 580 before being established that was not your established rating. The system treats everyone the same.
If you have others whose ratings are barely moving then their ratings are likely well established. Well-established players generally don't have a lot of movement unless they make specific efforts to improve or something happens which deteriorates their ability. Barely established players, of which you are currently, can have much more movement until there is a lot more games and opponents to validate against.
Mike has stated several times that game value is decreased over time.
it's true. as long as both players are rated correctly a 120 point difference means that the higher rated player will win most of the games played. Now that said the shorter the race the better chance to win the race. any given race to 3 can be won by the lower rated player but the majority of the races will be won by the higher rated player. Lenthen the race and the odds to win any given race drop considerably.
That is a great question. 200 games is where establishment happens which in statistics is referred to as confidence. 200 is where they say that they have a low confidence that the rating is likely to be correct enough to be established.
I personally would like to know what it looks like if two people had starter ratings a 100 points apart and then posted the exact same game results against the exact same skill levels. after 200 game would their ratings be exactly the same? If not why not?
yes
i get that no rating system is perfect, I’m only playing in one Fargo data capturing league & don’t play in tournaments. I’m not sure how the rating will ever even out if I stay in this league. When I play opponents that someone earlier reasonably suggested averaged out at 500, I was 80 points above them when I started. Now I’m 160 points above them but there’s no extra handicap, I’m still playing a race to one, not a race to 7 when I’d be giving something on the wire. Anyway I’m not trying to moan here, just trying to understand how this happened. I’d suggest not playing in a league doing races to one against lower ranked players if you’re trying to establish robustness 
Just curious, are you about even money to beat the 9 ball ghost on 5" factory GC pockets? A 660 is about even money, maybe a hair on the low side.
For the starter ratings assigned by some leagues, it was my understanding once 200 games were in the system, the starter rating influence would disappear entirely, and the fargo rating would take over 100%. And before 200 games were reached, a provisional rating would be some sort of averaging between the starter rating and the fargo rating for however many fargo games were played up until that point. Are both of these true?