hasu
Registered
In July 2023 I began tracking my Fargo performance rating on a night-by-night basis. If anyone's curious, there's a simple spreadsheet under "files" at the "Oklahoma Poolplayers" Facebook group called pr.xls that can be used to compute this. Equivalently, DigitalPool will compute your "Effective Rating" for an event that includes Fargo ratings.
This is something I started doing doing for fun, so that I could derive stats like how I tended to perform in league vs tournaments, how many games I tend to play a month, how many games it takes to move my rating, etc.
Something that has stood out to me is how much "stickier" my Fargo rating is than I expected. In a nutshell:
Weighted (by number of games) average performance rating from July 2023 - January 2024: 575
In other words, over my last ~500 games over a 6 month period, I've performed at the level of a 575, yet my rating has only increased by 24 points from 518 to 542. Not what I would have expected! It really goes to show how big of an influence older games have on your rating. It retrospect, this probably shouldn't have been so surprising. Taken from the Fargo's explanation of their ab initio global optimization process:
I've been playing for 7 years, and playing in Fargo-reported events for 6 years, so the player I was 3 years ago is significantly different from the player I am today. It's wild to imagine that those games still contribute half-weight to my rating! As someone who's still improving at a relatively steady rate, my guess is that a non-trivial "rating gap" will persist up and until I hit a wall in my progression and I perform at a relatively stable level for 12 months. But this is just a gut feel; Fargo's optimization process is non-linear, and my gut take is based on thinking about the data linearly.
Lest this be interpreted as complaining: I love Fargo and what's it's done to add objectivity to the sport. Also, how can I complain that I still get to play in "550 & under" tournaments
For anyone who's curious, here's my raw data:
There may be some minor inaccuracies in the raw data, as well as some oddities like (a) the fact that I played in a large BCA tournament in October that wasn't entered into the system until a month later, and (b) the fact that Fargo performs a nightly re-optimization. But the gross numbers like my weighted average performance should be quite accurate.
This is something I started doing doing for fun, so that I could derive stats like how I tended to perform in league vs tournaments, how many games I tend to play a month, how many games it takes to move my rating, etc.
Something that has stood out to me is how much "stickier" my Fargo rating is than I expected. In a nutshell:
| Rating | Robustness | |
|---|---|---|
| July 2023 | 518 | 1219 |
| January 2024 | 542 | 1724 |
Weighted (by number of games) average performance rating from July 2023 - January 2024: 575
In other words, over my last ~500 games over a 6 month period, I've performed at the level of a 575, yet my rating has only increased by 24 points from 518 to 542. Not what I would have expected! It really goes to show how big of an influence older games have on your rating. It retrospect, this probably shouldn't have been so surprising. Taken from the Fargo's explanation of their ab initio global optimization process:
I've been playing for 7 years, and playing in Fargo-reported events for 6 years, so the player I was 3 years ago is significantly different from the player I am today. It's wild to imagine that those games still contribute half-weight to my rating! As someone who's still improving at a relatively steady rate, my guess is that a non-trivial "rating gap" will persist up and until I hit a wall in my progression and I perform at a relatively stable level for 12 months. But this is just a gut feel; Fargo's optimization process is non-linear, and my gut take is based on thinking about the data linearly.
Lest this be interpreted as complaining: I love Fargo and what's it's done to add objectivity to the sport. Also, how can I complain that I still get to play in "550 & under" tournaments
For anyone who's curious, here's my raw data:
| Performance | Games played | Event | Date played | Rating | Rating change |
|---|---|---|---|---|---|
518 | |||||
744 | 9 | League | 13-Jul-23 | 520 | +2 |
685 | 9 | League | 20-Jul-23 | 522 | +2 |
572 | 9 | League | 3-Aug-23 | 522 | 0 |
541 | 9 | League | 10-Aug-23 | 522 | 0 |
553 | 9 | League | 17-Aug-23 | 522 | 0 |
542 | 31 | Tournament | 19-Aug-23 | 522 | 0 |
456 | 9 | Tournament | 20-Aug-23 | 522 | 0 |
630 | 9 | League | 24-Aug-23 | 523 | +1 |
544 | 9 | League | 31-Aug-23 | 524 | +1 |
784 | 9 | League | 7-Sep-23 | 527 | +3 |
555 | 9 | League | 21-Sep-23 | 527 | 0 |
501 | 9 | League | 28-Sep-23 | 528 | +1 |
541 | 9 | League | 19-Oct-23 | 528 | 0 |
587 | 25 | Tournament | 22-Oct-23 | 528 | 0 |
413 | 9 | League | 26-Oct-23 | 528 | 0 |
647 | 9 | League | 2-Nov-23 | 529 | +1 |
391 | 13 | Tournament | 4-Nov-23 | 529 | 0 |
472 | 9 | League | 9-Nov-23 | 528 | -1 |
648 | 33 | Tournament | 10-Nov-23 | 532 | +4 |
633 | 86 | Tournament | 11-Nov-23 | 541 | +9 |
598 | 25 | Tournament | 15-Oct-23 | 540 | 0 |
519 | 37 | Tournament | 15-Oct-23 | 540 | -1 |
491 | 9 | League | 16-Nov-23 | 540 | 0 |
625 | 9 | League | 30-Nov-23 | 540 | 0 |
511 | 27 | Tournament | 1-Dec-23 | 540 | 0 |
368 | 9 | League | 7-Dec-23 | 538 | -2 |
561 | 9 | League | 14-Dec-23 | 539 | +1 |
632 | 32 | Tournament | 17-Dec-23 | 541 | +2 |
647 | 5 | League | 4-Jan-24 | 542 | +1 |
481 | 5 | League | 11-Jan-24 | 542 | 0 |
539 | 15 | Tournament | 20-Jan-24 | 542 | 0 |
There may be some minor inaccuracies in the raw data, as well as some oddities like (a) the fact that I played in a large BCA tournament in October that wasn't entered into the system until a month later, and (b) the fact that Fargo performs a nightly re-optimization. But the gross numbers like my weighted average performance should be quite accurate.
