Can you let us know what tournament it was?
Fargo Rate - Not Much Math Here
- Thread starter BarTableMan
- Start date
But the "Bradley-Terry model" is not Fargo, "A player needs to have played a minimum of 200 games to have an "established" Fargo Rating. This is the standard "robustness" requirement, and without it, the rating is considered "preliminary".
Fargo is just an implementation of the Bradley-Terry statistical model for paired comparisons. The rest of your post seems like a mostly accurate but tautological description of how the model scores are shown to consumers.
Hitting a dictionary...
tautau wtf????? dude you gotta dumb that shit down brother. you are dealin with poolplayers.
I know plenty of APA 7/9's who can't run a rack in four tries, and most Fargo 300/350's I know have never run a rack. But I agree with the gist of your statement. When you start in any system, you go from "We know nothing about your play" to "This is all we know about your play". It takes a lot of play for any system to reliably say "This is how you play".
What’s the follow up from the OP?
Screenshot?
Actual rating?
Robustness?
Did someone assign you a starter rating (unlikely unless you played in a league match, rather than a one-off tournament)?
Is your ability such that you can regularly beat the ghost on soft equipment?
Screenshot?
Actual rating?
Robustness?
Did someone assign you a starter rating (unlikely unless you played in a league match, rather than a one-off tournament)?
Is your ability such that you can regularly beat the ghost on soft equipment?
I read it more as taking 4 innings to run a rack, not break and running 1 in 4 times.
For what it's worth,
--he's posting anonymously
--he's not responsive to follow-up questions
--I can't find a recent exchange in our support system that fits his description (might be missing it, though)
--None of the brand new players from the International under 650 event who started with no games currently has a preliminary rating over 650
I'm thinking we at this point have spent too much time on this.
Turns out I live in a rural area with a couple real small race to 2 or 3 weekly tournaments. I started getting them in the system a couple months ago. Every black-line rating here is a preliminary performance rating (no starter guess) based on an average of about 100 games.
Nobody but Vito here is established.
The light blue lines are the guesses I made before we started getting games in the system--just watching the players for 6 months or so.
I would say these preliminary ratings align pretty well with the subjective guesses. Michelle, with 83 games performing 79 points above my guess, is the farthest off.
Vito, by the way, is me: long story...
Edit.. I am replying to your reply which was not what I meant in my post LOL. You posted "run a rack in 4 tries" which seems to mean break and run, but I meant run a rack in 4 tries meaning 4 innings in a rack, not a 25% break and run rate.
There is no way a maxed-out APA player will not be able to run out a rack in under 4 innings most of the time with an open table. Not saying they will do so every single time, but most open table layouts for a 7/9 would be a one or two inning game, maybe 3 at times if it's not a 7-footer. It's just impossible to max out any league and not be able to run out well.
Last edited:
This is why I like Fargo, there is actual data instead of "I feel" or "I think" or "no way can..."
Mike is dumping games to take off the big 1000 entry payout
hahahah. <sarcasm!>
It would be interesting to see what their fargos would look like if you had used your guesses as starter ratings for them all.
I'm guessing it would be very similar, but I am still curious.
To the extent they have about 100 games, the preliminary ratings with starter-rating influence would be about halfway between the black and blue lines.
That's Vito who would enter THAT tournament ;-)Mike is dumping games to take off the big 1000 entry payout
Do you have stats on how “reliable” early results tend to be compared to established ratings? Like how 100 robustness for John Doe compares to 500 or 1000?
I’ve seen some examples of local people who were off by 50 or more points based on early results (and settle up or down to the correct level with more games), although by and large most seem to be about right (like maybe within 10 or 20 points of where they end up).
I also wonder whether the reliability of the low robustness results also can be influenced by low robustness of opponents.
Here is mine, just a sample size of 1. I kept track myself and made the graph. First picture is from 0 to 203 robustness. Second picture is from 0 to 1800 robustness (today).
