Are Fargo Ratings inflating?

[iusedtoberich]

I don’t think you can diagnose inflation/creep by looking at only one player, unless you know the results of all his games and the ratings of his opponents. He could have had a win rate that accounted for his 10 pt gain in 45 racks played.

That was just one example. If you look at the "same exact" top 10, top 20, etc players over the years, they almost all went up. I have posts bookmarked from 2015 showing the lists. Its night and day lower.

I personally think what has happened is as more data came in, it became clearer that the separation between the top pros and the rest of the pros was greater, and the separation between the rest of the pros and the bangers was greater, etc.

Here is one post from Mike in 2015:

US Open - Fargo Ratings of field

These are the Fargo Ratings of the 96 (of 128) participants for the US Open who have established ratings (at least 200 games in the system). These ratings are current as of 10-7-15 1 Shane Van Boening USA 824.5 2 Dennis Orcollo Philippines 811.5 3 Pin Yi Ko Taiwan, ROC 805.4 4 Niels Feijen...

forums.azbilliards.com

Here is another post from Mike in 2015:

Live Scoring and Brackets for US Open<<<<<<<<<<<<

This will be the link for live scoring and bracket info. It will be up and running when the time comes. SAVE it for your convenience. http://www.azbilliards.com/tours_and_events/158-us-open-9-ball-championship/brackets/

forums.azbilliards.com

Who wants to graph out the changes "per player" from these numbers to today?

I also had made a long term bet about Siming's rating in 2019, and for the purposes of the bet tracked the top 10 male pros every 6 months or so. The avg of all 10 top males strongly trended upwards.

 

[iusedtoberich]

Here is the "top 10 worldwide men" data. I was personally tracking this for a bet I made. Its the average fargo of whoever were the top 10 males at the date stamp.

 

[iusedtoberich]

Here are the names of those players in 2019 and in 2023. Many names are the same. We are to believe they all got better? I don't.

 

[iusedtoberich]

Now from the 2015 lists, the 2019 lists, the 2023 lists, and anyone can look up the 2024 lists, someone (not me) can make a graph of each player that is common. The graph will show older players like Shane and Neils will have gone way up. We all know Shane and Neils were better in 2015 than today.

Super old players like Efren and Busty have gone down, due to age. But they are harder to track because we don't have that data, only Mike does, or only someone like me who by chance took screenshots years ago randomly.

 

[iusedtoberich]

If anyone seriously wants to do it, I also have the top 10 male lists from between 2019 and 2023. I have maybe 10 time stamps saved, just like the post 23 above.

 

[skogstokig]

Now from the 2015 lists, the 2019 lists, the 2023 lists, and anyone can look up the 2024 lists, someone (not me) can make a graph of each player that is common. The graph will show older players like Shane and Neils will have gone way up. We all know Shane and Neils were better in 2015 than today.

Super old players like Efren and Busty have gone down, due to age. But they are harder to track because we don't have that data, only Mike does, or only someone like me who by chance took screenshots years ago randomly.

what about alex? agree on niels, pretty obvious

 

[Tom1234]

Very clear and self-explanatory, thank you.

The one thing that I am curious about: Fargo doesn't take into account the age of the player. I wonder what the average age of the players not improving during the 2800+ games played. I bet younger people who practice often are improving, and older folks are slowly declining, in spite of regular practice

I’ll bet the older folks declining would be a novel idea that the APA’s hcp system should take into consideration.

 

[iusedtoberich]

what about alex? agree on niels, pretty obvious

786 in 2015 per the link I shared a few posts above to a thread made by Mike.
811 today.

Any questions?

 

[Rocket354]

786 in 2015 per the link I shared a few posts above to a thread made by Mike.
811 today.

Any questions?

I agree with your points. Mike's narrative around his graph was not convincing to me.

Since Fargo us a relative measure, as more people enter the system and therefore more granularity is needed, the top players' ratings will necessarily inflate regardless of actual change in ability.

Fargo is a good cross-sectional measure, but not good for longitudinal studies, at least not without adjustments.

 

[mikepage]

[...]

In a stable state with no inflation, I think we’d see as many active players decreasing as increasing, so top players may increase 3.5 pts/year while bottom players decrease 3.5 pts/year.

In summary: to me the chart shows FR inflation of about 3.5 points per year.

There are 8354 players who had an established rating 2 years ago (average robustness 500) and logged between 1 and 100 games in the last two years (average 50). Forget the new games for a minute, and just think of the contribution to the rating from the original 500 games. If there was a general inflation of let's call it 3.5 points per year, then their opponents for those 500 games would have drifted up by 7 in those years and these players would be up 7 points even before accounting for the new play.

What we see, which is based on the original 500 games plus the 50 new games, is the group is up by 0.4 points (0.2 points per year). Further nearly half (46.4%) of the players went down in rating over the two years.

 

[mikepage]

[...]We all know Shane and Neils were better in 2015 than today.

Need to be careful to not conflate more dominant with better

 

[skogstokig]

786 in 2015 per the link I shared a few posts above to a thread made by Mike.
811 today.

Any questions?

yes. hohmann?

 

[garczar]

Need to be careful to not conflate more dominant with better

I was going to say that and i'm glad i waited til you added that. I believe your numbers/data pretty much are spot on. Some of it is kinda hard to wrap one's head around but its accurate.

 

[mikepage]

786 in 2015 per the link I shared a few posts above to a thread made by Mike.
811 today.

Any questions?

I have one. What if we look at players who were very active up to 2015 and then stopped playing (or at least logging games)?

If the general pool of their opponents has been riding a 3-point per year tide rise, then these inactive players should be up 25 points or something. Look at Chris Bartram --lots of games through about the end of 2014 and then stop. He was 752 in 2015 and is 754.6 now --up a few tenths of a point per year. Charlie Williams--same deal.

When you look for common players on top lists now and before, you introduce survivor bias.

Survivorship bias - Wikipedia

en.wikipedia.org

 

[derangedhermit]

Need to be careful to not conflate more dominant with better

Since Fargo is a relative rating, a rating based on play against others, what is the difference? Isn't dominance against opponents = better?

 

[DeadStick]

What we see, which is based on the original 500 games plus the 50 new games, is the group is up by 0.4 points (0.2 points per year). Further nearly half (46.4%) of the players went down in rating over the two years.

Thanks for the additional data, Mike.

So the least active of the active players went up by an avg of 0.2 pts per year, and the most active went up by 7 per year.

If the majority of the active players’ ratings went up over the last two years, how is that possible without general inflation?

 

[garczar]

Since Fargo is a relative rating, a rating based on play against others, what is the difference? Isn't dominance against opponents = better?

Dominance can also mean 'quality maintenance' in that the player maintained his/her level of play over time. Tiger didn't really get better at golf over time but he maintained his level over time. A really long time.

 

[mikepage]

Thanks for the additional data, Mike.

So the least active of the active players went up by an avg of 0.2 pts per year, and the most active went up by 7 per year.

If the majority of the active players’ ratings went up over the last two years, how is that possible without general inflation?

There are many more players in the first category. If you look at the plot in the first post, the dots representing a lot of play each are an average of 1,000 players. The one dot alone representing the least play is an average over 12,000 players. Plus there are many players who haven't logged new games at all.

 

[mikepage]

Since Fargo is a relative rating, a rating based on play against others, what is the difference? Isn't dominance against opponents = better?

That's the distinction I'm warning about, between better-relative and better-absolute.

 

[BasementDweller]

I won't pretend to know what's going on but I'm quite curious.

2015 Corey Deuel: 762
2024 Corey Deuel: 775

My eyes tell me he hasn't gotten better in the past ten years but relative to the top players I can intuit 775 as being correct. What gives if he hasn't gotten better? Could it be a rising tide lifting all boats situation, where the very top tier of players have gotten so good that they routinely clean up against lesser competition? More so than they did just a few short years ago. Sort of like what Filler has done this weekend (9-0, 9-0). Could there be a "Fargo" effect, where players are grinding harder when they know their Rating is at play? My eyes tell me the top players have gotten better and are delivering the performances to show it. Could this increased performance at the top be driving what looks like inflation?