Can I question my Fargo rating?

[MattPoland]

I think you should consider it’s not just about the average of your opponents. It’s which opponents you beat and which you lose to. I win 70% of my matches but I’m 557. Then again I think my average opponent may be a lower than 500.

But those win% by rating bars tell a lot about how your wins are distributing. I have trouble with a few specific weaker players that are great shot makers but can’t control the cueball or work patterns. But for some reason if I get deep on them and don’t get out, any time they miss they seem to always leave me hooked.

In terms of league play, there is one tiny asterisk. I usually shoot anchor so I’m in a position where if I need 5 balls, I might lose the game not because of sandbagging but because I’m prioritizing getting five ducks off the table before addressing problems. Where if I need to win and keep my opponent to 1-3 balls then I’m likely going aggressive for a TR if I don’t see any lock-up safeties along the way. But if all I need is a win, I’m likely to play my normal tournament/action game doing early breakouts, run if I can, safety if I can’t, tie up balls, whatever is needed to make it happen

 

[Bob Jewett]

Bob -- don't both the methods that you describe here actually rely on the same sort of system, which is relative records of players against other players? ...

No. As IUTBR pointed out above, sometimes top players don't want to enter tournaments or can't.

A different kind of anomaly was in the carom organization where ranking was based on participation largely. One guy played in every event and accumulated a lot of points one or two at a time. He ended up highly ranked just because of perfect attendance.

 

[mikepage]

No. As IUTBR pointed out above, sometimes top players don't want to enter tournaments or can't.

A different kind of anomaly was in the carom organization where ranking was based on participation largely. One guy played in every event and accumulated a lot of points one or two at a time. He ended up highly ranked just because of perfect attendance.

The Eurotour is an extreme of this. One guy was ranked top 100 with a Fargo Rating in the 300s and having never gotten to 5 in a race to 9 after many many tournaments. Many others rated over 700 ranked higher than 100.

 

[Bob Jewett]

The Eurotour is an extreme of this. One guy was ranked top 100 with a Fargo Rating in the 300s and having never gotten to 5 in a race to 9 after many many tournaments. Many others rated over 700 ranked higher than 100.

Persistence and grit are rewarded.

It's good to know that I have some chance against at least one Eurotour regular. It's just not Josh.

 

[mikepage]

I think you should consider it’s not just about the average of your opponents. It’s which opponents you beat and which you lose to. I win 70% of my matches but I’m 557. Then again I think my average opponent may be a lower than 500.

But those win% by rating bars tell a lot about how your wins are distributing. I have trouble with a few specific weaker players that are great shot makers but can’t control the cueball or work patterns. But for some reason if I get deep on them and don’t get out, any time they miss they seem to always leave me hooked.

I think this is more perception that reality. You have, as you know, a lot of games against Bobby D., rated 494. Your performance against players stronger than Bobby D (89 games) and your performance against players weaker that Bobby D (160 games) are the same within 2 points.

In terms of league play, there is one tiny asterisk. I usually shoot anchor so I’m in a position where if I need 5 balls, I might lose the game not because of sandbagging but because I’m prioritizing getting five ducks off the table before addressing problems. Where if I need to win and keep my opponent to 1-3 balls then I’m likely going aggressive for a TR if I don’t see any lock-up safeties along the way. But if all I need is a win, I’m likely to play my normal tournament/action game doing early breakouts, run if I can, sa

The size of this asterisk is indeed small. Before accepting 8-Ball games using the 10-point system, we analyzed many tens of thousands of these games to make sure the occasional incentive issue (need a few points to win the round so will lower chance of winning the game to go after them) was not significant. It isn't. You may occasionally pocket a breakout ball early because you need the point. But in this league you also occasionally benefit from an opponent pocketing a breakout ball early. We don't detect a difference after a lot of games.

 

[MattPoland]

I think this is more perception that reality. You have, as you know, a lot of games against Bobby D., rated 494. Your performance against players stronger than Bobby D (89 games) and your performance against players weaker that Bobby D (160 games) are the same within 2 points.



The size of this asterisk is indeed small. Before accepting 8-Ball games using the 10-point system, we analyzed many tens of thousands of these games to make sure the occasional incentive issue (need a few points to win the round so will lower chance of winning the game to go after them) was not significant. It isn't. You may occasionally pocket a breakout ball early because you need the point. But in this league you also occasionally benefit from an opponent pocketing a breakout ball early. We don't detect a difference after a lot of games.

I’m certainly not calling foul on my rating. I just think when it comes to think about win rates apples to average ratings, it helps to know you’re still expected to have a bigger win rate against opponents below that average than of opponents above that average. Obviously lower robustness has bigger sways but ultimately even a little drift up or down is going to happen when your performance doesn’t match expectations of your rating (up or down).

But maybe those losses to Debbie, Kay, Cindy, and Loni still haunt me because they are “rolling cueball only” players. But my long term results against each of them probably stacks up to expectations. Hard not to dwell on how those individual games went down.

 

[peppersauce]

Because I can perform better in the format then compared to tournament play. If stronger results don't equate to a stronger rating, then I'm really mis-understanding the concept behind fargo.

We put a team of all 600+ players together and played a couple of sessions. Nobody’s rating really changed much at all, despite the fact that we were absolutely smashing the other teams, if you look at games won. (Handicap—that’s another story) Both sessions, 3-4 out of 5 of our core players were in the top 10 by win percentage at the end of the session. The lowest rated player was just over 600 and the highest was in the 660s. I think my win percentage the last session was something like 85%. My rating barely moved, and there were still a handful of 600+ players NOT on our team. Highest rated player those two sessions was a 727 on another team.

It might seem like you perform better, but it might just look that way in a race to 1 format against so many different skill level players.

 

[tomatoshooter]

I think you should consider it’s not just about the average of your opponents. It’s which opponents you beat and which you lose to. I win 70% of my matches but I’m 557. Then again I think my average opponent may be a lower than 500.

But those win% by rating bars tell a lot about how your wins are distributing. I have trouble with a few specific weaker players that are great shot makers but can’t control the cueball or work patterns. But for some reason if I get deep on them and don’t get out, any time they miss they seem to always leave me hooked.

In terms of league play, there is one tiny asterisk. I usually shoot anchor so I’m in a position where if I need 5 balls, I might lose the game not because of sandbagging but because I’m prioritizing getting five ducks off the table before addressing problems. Where if I need to win and keep my opponent to 1-3 balls then I’m likely going aggressive for a TR if I don’t see any lock-up safeties along the way. But if all I need is a win, I’m likely to play my normal tournament/action game doing early breakouts, run if I can, safety if I can’t, tie up balls, whatever is needed to make it happen View attachment 709200

That's pretty funny how your win percentage is slightly worse against <300 players. Somehow they always seem to leave the cue ball in the worst place, and create clusters. Do you know if the 300+ is just the band from 300-400, or is does it include 400s, 500s, etc.?

 

[hang-the-9]

No. As IUTBR pointed out above, sometimes top players don't want to enter tournaments or can't.

A different kind of anomaly was in the carom organization where ranking was based on participation largely. One guy played in every event and accumulated a lot of points one or two at a time. He ended up highly ranked just because of perfect attendance.

This is why any of the tour or organization ratings need to be looked at not for the exact pecking order of the players overall but only for that specific event or series. We all know the several lists of top players looked different depending on who did the list. That is why I go with Fargo ratings, it just tells us who beat who and by how much, now how often they played.

 

[The_JV]

We put a team of all 600+ players together and played a couple of sessions. Nobody’s rating really changed much at all, despite the fact that we were absolutely smashing the other teams, if you look at games won. (Handicap—that’s another story) Both sessions, 3-4 out of 5 of our core players were in the top 10 by win percentage at the end of the session. The lowest rated player was just over 600 and the highest was in the 660s. I think my win percentage the last session was something like 85%. My rating barely moved, and there were still a handful of 600+ players NOT on our team. Highest rated player those two sessions was a 727 on another team.

It might seem like you perform better, but it might just look that way in a race to 1 format against so many different skill level players.

The bolded is the key point. A change in rating regardless of direction would be dependent on the average quality of the opponents.

This is probably my sense of reality clashing with statistics logic once again.

 

[JB Cases]

Once again, thanks for all the input everyone has been offering. It’s been great. A decade ago I used to play a lot of pool, now i play less, am older & my general interest in it has dropped. I liked the thought that “maybe I play better than I think I do”. That’s a very reasonable question. There’s some truth to that but not enough, even though I am my own worst critic. I do know I miss more regularly than I used to. The players in the league with similar winning %s to me, and who I’d rate as similar speed, are 40 to 50 points below me. The league is competitive, everyone in it seems to try to win to the best of their talents. I guess I have to hope that they start beating me more frequently

One thing I have found with Fargo ratings is that they tend to tell truth to ego by effectively showing where a player REALLY is in the pecking order of skill. For example I would have rated myself as closer to pro level than I am despite being in full knowledge that I was still far below them. Now I know exactly how far below them I am and due to Fargo Ratings I can correlate the skill being displayed with a number and compare that observation to others and their numbers. Thus I know the types of mistakes that are often made by those at my level compared to the mistakes made by those above and below me. This is both a blessing and a curse in that it helps to match up fairly but also shows how far one is away from where they thought they actually were.

 

[mikepage]

One thing I have found with Fargo ratings is that they tend to tell truth to ego by effectively showing where a player REALLY is in the pecking order of skill. For example I would have rated myself as closer to pro level than I am despite being in full knowledge that I was still far below them. Now I know exactly how far below them I am and due to Fargo Ratings [...]

Sorry Buddy...

 

[JB Cases]

No rating system can account properly for players that don't want to play to their ability. ANY system. If I am the best basketball player, but when I go play in front of a crowd I miss shots on purpose, I am not going to be drafted. The only rating system that can work this way is the guy in the corner watching the players for years that has developed a local knowledge of the players personally and knows what they can do when they are trying.

Just like Scarlet, rating systems depends on the kindness of strangers, although in their case it's the "honesty" of strangers that is needed. Even for "fun" games, if they are reported as results that affect the handicap, taking it easy on a buddy will mess up the data. No way around it. Have fun and let the people win when not in an official match. The good thing about Fargo is that it does not go up and down like many other league systems, it takes quite a bit of losing or winning to move the points enough to matter. I have been a 550ish rating for years, with maybe getting to 555 or so once in a while, but always 550-555, over years of playing. That is a less than a 1% change. It would need to be about 10% or higher for someone to notice a marked difference in skill over a period of time, and moving 10% with an established rating takes some work, more so than a few dumped games.

This is correct that any system that depends on user-generated data which is partially or in full controlled by the user can have inaccurate results based on flawed input. Each player essentially has one thing almost fully under their control and that is the choice to lose on purpose. I say almost because if the opponent is also intent on losing deliberately then they might get there first.

So the only way short of outright fraudulent reporting of games that weren't even played to influence your rating is by losing games constantly against any opponent you play. Enter tournaments with the longest races and go two and out long enough and the system will reflect a pretty low rating for you.

The thing is though is that human ego and human laziness often prevent most people from developing the desire to lose on purpose and the laziness prevents them from doing the work to figure out the algorithm sufficiently to bypass any safeguards built in to recognize attempts to game the system.

regarding the difference between dinking off games in "fun" leagues while being serious in "cash" leagues......well, that level of effort is also part of who a player is and it works it's way into all matches played because in my opinion there are very few people who possess the ability to truly focus or not like turning a switch on or off. And even when they think that they can the added "must win" pressure likely causes them to dog shots that they wouldn't dog for fun. This is actually a well known phenomena among players in reverse when a player is known as a champion for free and a dog for the cash. I think that in general the normal players, not the card-carrying hustlers, have less ability to turn the focus on and off than they think that they do. The hustlers have developed the ability as a part of their job skills.

And you're right that the steermen who kept their fingers on the pulse of all the players around them had the ability to confidently rate players comparatively. The difference though is that when a player has been out of view for a while then their data is old. What Fargo doesn't see and which I wish there was a way to at least report on through trusted sources is the money matches. Some players have a fargo of say 550 but they regularly gamble with players well above them and win in matches that are not captured in the rating system. So there is a disparity in the actual skill level rating and the actual skill level performance. I don't know how that can be reconciled for the few that have that disparity.

Fargo needs to have some trusted reporting access for those in positions of observation such as tournament directors and room owners. Although even then someone would likely abuse that trust just to provide fraudulent information.

 

[JB Cases]

Sorry Buddy...

I am ok with it. Because I can also correlate it with the effort to improve by watching ratings go up as skills are acquired and sharpened. Fargo gives goals that are concrete and not nebulous.

 

[JB Cases]

I’m showing my general pool ignorance here by telling you I’m not even sure what the ghost is about.

for a casual player that come in at 660 and not to know what the ghost is I would be pretty proud of that. You're an above average player for sure!

 

[BasementDweller]

I reject the underlying assumption being presented here that -- effort level can't skew the numbers.

In what other sport would this not be true? Go to the gym to play pickup basketball, only to find the court full of new teenage players. If you play, are you going to go all out?

Likewise in a social league, where you're playing say - an outmatched team of flirty females, are you going to destroy them?

Maybe only 1 person in the thread has access to the data, but surely more than that have some common sense.

Now is this a HUGE problem? That's doubtful, especially as more people take pride in their ratings. I tend to believe true competition ultimately brings out the best in people -- in the long run. A sort of rising tide lifting all boats sort of deal.

Humans aren't machines, even if the data makes it look like we are.

 

[OverTheNut]

Some good info in this thread. As someone who has recently started playing in a Fargo reported league I did wonder about a couple of things.

Mainly - what if someone starts at 325 and someone else at 525 even though they are about the same skill level and post up similar results. The answer seems to be that after about 200 games they'd probably have about the same Fargo rating - regardless of their starting number.

This is obviously a good thing. Let the algorithm do it's thing. It's pretty tried and tested.

 

[iusedtoberich]

Now that this thread has been bumped 7 months later, what is your current rating and robustness, @roger.p?

 

[iusedtoberich]

Some good info in this thread. As someone who has recently started playing in a Fargo reported league I did wonder about a couple of things.

Mainly - what if someone starts at 325 and someone else at 525 even though they are about the same skill level and post up similar results. The answer seems to be that after about 200 games they'd probably have about the same Fargo rating - regardless of their starting number.

This is obviously a good thing. Let the algorithm do it's thing. It's pretty tried and tested.

Yes, if you were assigned a "starter rating" its influence completely goes away at exactly 200 games of robustness. Most players are NOT assigned a starter rating. Additionally, if you were assigned a starter rating, and it turns out to be way off, it can be adjusted closer to your "preliminary" calculated fargorate (when under 200 games) by the league operator. It can NOT be adjusted further away by the league operator.

When you look up someone's rating in the app, AND they have a P next to their name indicating under 200 robustness, it will show if they have a starter rating or not. 0 means no starter rating. Any number besides 0 means a league operator's guess.

 

[Chili Palmer]

Yes, if you were assigned a "starter rating" its influence completely goes away at exactly 200 games of robustness. Most players are NOT assigned a starter rating. Additionally, if you were assigned a starter rating, and it turns out to be way off, it can be adjusted closer to your "preliminary" calculated fargorate (when under 200 games) by the league operator. It can NOT be adjusted further away by the league operator.

When you look up someone's rating in the app, AND they have a P next to their name indicating under 200 robustness, it will show if they have a starter rating or not. 0 means no starter rating. Any number means a league operator's guess.

Confirmed. I don't remember my starter rating but I do know that at one point I was playing more than one standard deviation better than my rating (i.e. I was winning 80% of my games when I should've been winning 60% of my games - I don't remember the ACTUAL numbers, I just know it was more than one standard deviation) so I was bumped up 100 points. I think I was given 350 for starter, after a couple of months I was rated 450'sh and was playing at around 550'sh so I was bumped up to 550'sh. Quit playing a couple of years ago with just over 200 robustness and just over 600 FR. Again, I don't remember specifics but I do remember being informed by Mike Page of the change.