Fargo Breakdown - Singles @ BCAPL Nationals

[watchez]

CSI STATEMENT - The Rationale Behind Division Boundaries

Please allow us to explain the rationale for the division boundaries this year.

Deciding where the singles divisions should be split involves two main considerations that need to be balanced, a) the rating difference from top to bottom and b) the size of the fields. The goal is to balance these two aspects in the best possible way. However, this first year of using FargoRate has a third variable that must be considered – players with no data and only a starter rating. Many players have no data at all and come into the event with only a starter rating (i.e. 625, 525, 425, etc.) based on their previous classification. History has shown that these starter ratings are often lower than the person’s actual ability.

This sentence here in red is conflicting to me. CSI is the organization that gave these players their starter rating of 625,525,425 etc. It isn't a FargoRate determination. So, CSI feels that these starter ratings are "often lower than the person's actual ability", why didn't you simply have a starter rating that put these unknown, unranked players higher.? You could have started them at 650,550,450,etc. It was CSI choice to make them what they are to start.

At least CSI is stating that they don't feel that 80% of these unranked players that are 525 are going to actually be lower when they have a FargoRating, as someone tried to infer previously.

 

[robsnotes4u]

This sentence here in red is conflicting to me. CSI is the organization that gave these players their starter rating of 625,525,425 etc. It isn't a FargoRate determination. So, CSI feels that these starter ratings are "often lower than the person's actual ability", why didn't you simply have a starter rating that put these unknown, unranked players higher.? You could have started them at 650,550,450,etc. It was CSI choice to make them what they are to start.



At least CSI is stating that they don't feel that 80% of these unranked players that are 525 are going to actually be lower when they have a FargoRating, as someone tried to infer previously.

What are you trying to prove?


Mike just did an exercise, pulling data, to show you and others that it works. He backs up his statements with data. Instead of saying thank you, or that makes sense, you switch the discussion to something else.

Why not be a part of making pool better?



Sent from my iPhone using Tapatalk

 

[bad_hit]

We had "Open" (yeah right), "Advanced", "Master"..."A", "B", "C"....

Everything is the same except now we have #s that, with enough data, correlate a little better to someone's skill level.

Perfect? No. Better? Of course it is.

C'mon...Open? It was a free for all. Didn't Shane win that one year?

 

[JC]

It's nice to know what a fair race is statistically but what kind of a nit wants to play tournament play accepting two on the wire and then pounding their chest as a "winner" against someone they are not qualified to rack for? (Don't answer that, I already know these people, they are called "APA")

I believe the real value is in grouping players of like skills to have a fair event, not in handicapping an event open to all in one division based on odds of winning handicapped.

Which of course is what CSI is doing and good for them!

JC

 

[watchez]

This is not a small piece; it is core to what we do.

I don't make claims about exact numbers. But within expected statistical fluctuations, yes.

85 points is 85 points

Below I show all matches we have between SVB and opponents around 85 points below him. I go up 7.5 points in both directions to get better statistics. There are 601 games, and Shane is supposed to win in a ratio of 1.80 to 1.

That ratio is Shane 386 games, opponents 215 games

The actual record is Shane 391, opponents 210

In other words, this group of players is performing at a level 85 points below Shane, as expected. And yes, we see this same kind of agreement when we look at 85 points for low-level league players. That is key to what we do.

The matches here are Shane's score, opponent's score

Mike - can you do this for the guy that won the BCAPL Open singles last year? I believe his name was Jon Brown

 

[sbpoolleague]

CSI STATEMENT - The Rationale Behind Division Boundaries

Thank you for the clarifying statement, CSI. We will see you in 8 days.

 

[robsnotes4u]

Mike - can you do this for the guy that won the BCAPL Open singles last year? I believe his name was Jon Brown

Jon use to live in Fargo. Has a daughter in ND and plays there every once in awhile. I played him in a tourney in Moorhead, across the river last year.

Here is a little from that tourney. Jon versus Randy Hanson. I think this is when I had the the iPad flipped so it looks like they are shooting opposite handed.

http://ustre.am/M6h4



Sent from my iPhone using Tapatalk

 

[sbpoolleague]

Mike - can you do this for the guy that won the BCAPL Open singles last year? I believe his name was Jon Brown

Here is Jon Brown and his opponents from the Men's Open Singles last year. Note that Brown and six of his opponents (all of which finished in the top 25) would not be eligible for the Gold division.

Jon Brown - Fargo Rating = 733 (178)

Losing
Score Opponent Fargo (Robustness)
-----------------------------------
2 - Tony Jacobs 569 (311)
0 - Colin Angle 521 (13)
X - Steve Register 502 (674)
1 - Emil Richards 560 (41)
1 - Alex Kanapilly 614 (312)
2 - Cory Livermore 661 (161)
3 - Jarred Schlauch 675 (384)
1 - Jerrod Frideres 679 (956)
4 - Ryan Robinson 687 (333)
2 - Darcy Gilkes 661 (805)
2 - Lance Schofield 702 (528)

 

[sbpoolleague]

Not sure if this analysis has been done before, but anyways...

Here is a list of the top 48 players in last year's Men's Open Singles.

You will note that 44 of the 48 players would not even be eligible for this year's Gold division, and the winner, Jon Brown, would not be eligible for the Platinum division.

Almost 96% of the top-48 money went to those players ($58,800 of $61,400).

 

[watchez]

How many of those 48 are signed up for singles this year?

 

[watchez]

Here is Jon Brown and his opponents from the Men's Open Singles last year. Note that Brown and six of his opponents (all of which finished in the top 25) would not be eligible for the Gold division.

Jon Brown - Fargo Rating = 733 (178)

Losing
Score Opponent Fargo (Robustness)
-----------------------------------
2 - Tony Jacobs 569 (311)
0 - Colin Angle 521 (13)
X - Steve Register 502 (674)
1 - Emil Richards 560 (41)
1 - Alex Kanapilly 614 (312)
2 - Cory Livermore 661 (161)
3 - Jarred Schlauch 675 (384)
1 - Jerrod Frideres 679 (956)
4 - Ryan Robinson 687 (333)
2 - Darcy Gilkes 661 (805)
2 - Lance Schofield 702 (528)

I wanted to see Jon Brown vs people that are 85 pts less than him -- just like the SVB data

 

[robsnotes4u]

I wanted to see Jon Brown vs people that are 85 pts less than him -- just like the SVB data

I doubt you would get a true picture with his I established rating and a lot of those games being in that tournament.


Sent from my iPhone using Tapatalk

 

[Jeff G. Martin]

Not sure if this analysis has been done before, but anyways...

Here is a list of the top 48 players in last year's Men's Open Singles.

You will note that 44 of the 48 players would not even be eligible for this year's Gold division, and the winner, Jon Brown, would not be eligible for the Platinum division.

Almost 96% of the top-48 money went to those players ($58,800 of $61,400).

Curious why you didn't include each player's Robustness in this chart like you did above in the Jon Brown match breakdown. I noticed a lot of players are very Preliminary and as such could have wound up in Gold with their Starter Rating and taken the money anyways.

 

[watchez]

I doubt you would get a true picture with his I established rating and a lot of those games being in that tournament.


Sent from my iPhone using Tapatalk

That's strange. I read where you and others said at 200 games your Fargo rating is valid. Jon is at 178.....that's approximately 200. See what I did there. And why would 'that tournament' matter? I've been told all data is good data. Shouldn't matter where it came from. In fact a national tournament where everyone is playing for a substantial prize and trying their best would be better than some Tuesday night $10 entry tournament race to 3. Or are you saying the Fargo Rate is only valid when you want it to be?

 

[robsnotes4u]

Do you really even read what you write?

The data is good from the games in the tourney is good.

The data you are looking for would be from games when he isn't established. Would his rating mean more at 200 games when it is established or 120 when it is not.

The tourney he is playing in doesn't matter, a game is a game.

You are falling deeper into the rabbit hole.

You must follow Eberle and his flat world idea.

Let us know how you feel after nationals.




Sent from my iPhone using Tapatalk

 

[jojopiff]

Not a FargoRate issue but I run a little league with around 50 players. I dealt with a Lucy in the BCA office and asked what she needed & I was told only wins/losses for each player. Not broken down as to who beat who, only that John Doe went 65-20. Well I know this isn't the way FargoRate is going to work but I'm not going to beg someone else to do their job. Sure enough, I've had several people asking me why they don't have a rating. Some are irrelevant but one guy is gonna be a 525 when, in fact, he should probably be the next level up. He emailed me tonight about the fact he doesn't have a FargoRate & I said "I'm not in charge of your FargoRate, all I do is turn the info BCA asked for in to them. But be happy because you're very under-ranked." The whole system can only be as good as the people you employ. I can't imagine I'm the only person who runs a league has run into this.

 

[jojopiff]

Good, then stop right there and go with the flow. Arguing ad nauseum about a topic that you know nothing about is a huge waste of time and it's confusing to all the others that know equally little.

Again, I have this stupid thing I do called "thinking" and don't walk over a cliff to my death cause that's what everyone is doing. You do you, I'll do me. If you have a problem with that then go f... Well I think you know the end of that.

 

[jojopiff]

This is not a small piece; it is core to what we do.

I don't make claims about exact numbers. But within expected statistical fluctuations, yes.

85 points is 85 points

Below I show all matches we have between SVB and opponents around 85 points below him. I go up 7.5 points in both directions to get better statistics. There are 601 games, and Shane is supposed to win in a ratio of 1.80 to 1.

That ratio is Shane 386 games, opponents 215 games

The actual record is Shane 391, opponents 210

In other words, this group of players is performing at a level 85 points below Shane, as expected. And yes, we see this same kind of agreement when we look at 85 points for low-level league players. That is key to what we do.

The matches here are Shane's score, opponent's score

Appreciate the up front response and data. Very good of you to provide to show transparency. However, you stated before it's a 56/44 ratio whilst SVB shows over a 64% win ratio. Could you show the rest of the people who are in about the 50% ratio bringing this down to the 56/44 overall ratio? For people of 85 point differences.

 

[mikepage]

Not a FargoRate issue but I run a little league with around 50 players. I dealt with a Lucy in the BCA office and asked what she needed & I was told only wins/losses for each player. Not broken down as to who beat who, only that John Doe went 65-20. Well I know this isn't the way FargoRate is going to work but I'm not going to beg someone else to do their job. Sure enough, I've had several people asking me why they don't have a rating. Some are irrelevant but one guy is gonna be a 525 when, in fact, he should probably be the next level up. He emailed me tonight about the fact he doesn't have a FargoRate & I said "I'm not in charge of your FargoRate, all I do is turn the info BCA asked for in to them. But be happy because you're very under-ranked." The whole system can only be as good as the people you employ. I can't imagine I'm the only person who runs a league has run into this.

That's not on Lucy nor on CSI. That's on us. What's required is league-management software that feeds data in, and we are testing the solution right now.

 

[mikepage]

Appreciate the up front response and data. Very good of you to provide to show transparency. However, you stated before it's a 56/44 ratio whilst SVB shows over a 64% win ratio. Could you show the rest of the people who are in about the 50% ratio bringing this down to the 56/44 overall ratio? For people of 85 point differences.

Sorry, that 44% was an error.

With an 85-point gap, the lower rated player is expected to win 35.7% of the games
higher rated player, 64.3% of the games

And it is this (64.3/35.7) that is 1.8-to-1

The 35.3% is simply

1/(1+2^(ratinggap/100)) , with ratinggap=85 in this case