I've been glad for many years you once enjoyed talking with the guys from the block plant. My brother-in-law worked there until he passed.And just to clarify, if your ranking isn’t what you think it should be, it is 100% because I hate your team.
Testing the Toy
- Thread starter 51eleven
- Start date
granger
Founder
When I calculate error, I cap any win over 45 at 45. As long as the expected was 45 or over, the error is 0. This is what makes things tough. You cannot use a true error. If a team was supposed to win by 100 and wins by 45, that's really all you can do.
As for @Dilla Killa asking about 1-x sims or regressions, no, I do not. There is literally only one input -- SCORE. You cannot get anything else. Heck, it is hard enough just getting the scores. So, this becomes really more rules-based than some sort of deep learning.
Remember, I started this in 1993 and there wasn't enough power on a laptop to run anything too serious at that point and you still only have the one input. So maybe it is best to dive into the evolution of the whole thing.
The idea came from when a college roommate who played chess taught me about the ELO rating system. Two people play chess and are rated X and Y. Based on the difference, X should win z% of games. So, based on what actually happens, you follow a formula and adjust each player. At that time Sagarin was the only person doing ratings (in USA Today) and I followed those closely.
So I though, there has to be an IDEAL team (not real, but fictional perfect team). Let's make that team 100. Everyone else is somewhere between that perfect 100 and 0. This was all completely arbitrary. Next what do you do with the score. Well, I had collected a ton of scores and calculated that the average teams wins by x with a standard deviation of y. I don't remember the actual numbers, but every year in the early years, I would make sure that stayed close to true and adjust accordingly. Now is Team A beats Team B by let's say 24 and they were expected to win by 11, they over-performed by 13 and the other team under-performed by 13. You get the t-score of that and multiply it by the average MOV and you move each team either +/- that.
Interesting side note - the original algorithm was written in Pascal on an Apple. That went away fairly quickly and I had to move to straight C because I got a work PC and my Apple Pascal no longer worked. Then C++ then MS Basic. The last version I wrote was in Python.
OK, back to the early days. It seemed to work fine for small groups of schools, but the sport started growing and you had to go beyond 100, especially to keep teams from gradually falling below 0. I never wanted a team to have a negative ranking. I always believed in 2 things: every team deserved to be ranked and no team should fall below 0. It was important for me to care about everyone (except for your school, which I obviously hate
)
As the system grew, so did the need to enable larger moves. The concept of upsets was added as well as the idea that beating a team ranked in the top-10 or 20 should maybe provide a bump of some sort. This all came together with the idea of continually re-running the system in micro bursts until it reached some sort of equilibrium. Over the years I have tried different ways of scaling this. None are perfect.
In the early years, I never worried about the error much and would wait until the season was over to tweak. Later, I would sometime tweak a little and test for the first few weeks.
About 10 years ago I took the errors to heart and started tracking the different ones, but it came down to what is most important? I went with PREDICTING THE WINNER narrowly over MARGIN OF VICTORY, so I try to incorporate both, but in my mind, picking winners is the more important thing.
As for @Dilla Killa asking about 1-x sims or regressions, no, I do not. There is literally only one input -- SCORE. You cannot get anything else. Heck, it is hard enough just getting the scores. So, this becomes really more rules-based than some sort of deep learning.
Remember, I started this in 1993 and there wasn't enough power on a laptop to run anything too serious at that point and you still only have the one input. So maybe it is best to dive into the evolution of the whole thing.
The idea came from when a college roommate who played chess taught me about the ELO rating system. Two people play chess and are rated X and Y. Based on the difference, X should win z% of games. So, based on what actually happens, you follow a formula and adjust each player. At that time Sagarin was the only person doing ratings (in USA Today) and I followed those closely.
So I though, there has to be an IDEAL team (not real, but fictional perfect team). Let's make that team 100. Everyone else is somewhere between that perfect 100 and 0. This was all completely arbitrary. Next what do you do with the score. Well, I had collected a ton of scores and calculated that the average teams wins by x with a standard deviation of y. I don't remember the actual numbers, but every year in the early years, I would make sure that stayed close to true and adjust accordingly. Now is Team A beats Team B by let's say 24 and they were expected to win by 11, they over-performed by 13 and the other team under-performed by 13. You get the t-score of that and multiply it by the average MOV and you move each team either +/- that.
Interesting side note - the original algorithm was written in Pascal on an Apple. That went away fairly quickly and I had to move to straight C because I got a work PC and my Apple Pascal no longer worked. Then C++ then MS Basic. The last version I wrote was in Python.
OK, back to the early days. It seemed to work fine for small groups of schools, but the sport started growing and you had to go beyond 100, especially to keep teams from gradually falling below 0. I never wanted a team to have a negative ranking. I always believed in 2 things: every team deserved to be ranked and no team should fall below 0. It was important for me to care about everyone (except for your school, which I obviously hate
)As the system grew, so did the need to enable larger moves. The concept of upsets was added as well as the idea that beating a team ranked in the top-10 or 20 should maybe provide a bump of some sort. This all came together with the idea of continually re-running the system in micro bursts until it reached some sort of equilibrium. Over the years I have tried different ways of scaling this. None are perfect.
In the early years, I never worried about the error much and would wait until the season was over to tweak. Later, I would sometime tweak a little and test for the first few weeks.
About 10 years ago I took the errors to heart and started tracking the different ones, but it came down to what is most important? I went with PREDICTING THE WINNER narrowly over MARGIN OF VICTORY, so I try to incorporate both, but in my mind, picking winners is the more important thing.
It was really nice to see the Toy, which I respect and love be so wrong tonight. By 43 points. Almost 49. Though I saw the Blum scrimmage, my first live game of the year. With an old friend/co-worker I'd not seen in about 37 (?) years, who had never seen a sixman game. He loved it.
This is not worth a thread of its own, so I'll report it here.
Strawn went down 24 -8 in the first quarter. Then by only 8 at the half. Down by 24 again in the third. Then with about 2 minutes left 52-50, a missed xp would have tied it. Gorman fumbles. Last play. 20 yard pass to receiver at the 5 near the sideline. Reverses field, goes backward and outraces the defense to the two on the other sideline before being tackled. Fun game to watch in one of the coolest stadiums in sixman. I think this group of Hounds maybe grew up a little tonight.
This is not worth a thread of its own, so I'll report it here.
Strawn went down 24 -8 in the first quarter. Then by only 8 at the half. Down by 24 again in the third. Then with about 2 minutes left 52-50, a missed xp would have tied it. Gorman fumbles. Last play. 20 yard pass to receiver at the 5 near the sideline. Reverses field, goes backward and outraces the defense to the two on the other sideline before being tackled. Fun game to watch in one of the coolest stadiums in sixman. I think this group of Hounds maybe grew up a little tonight.
Dilla Killa
Go Coyotes!
Understood. In the early 90's, we were working on similar issues as I mentioned above. Back then, we had Cray's to crunch the various physics related algos. Many, many millions of dollars for the computing power to do only a fraction of what we can do now with a 16 core Xeon for $18k. Good times.
justlookin'
Member
The feeling is mutual old friend. Miss seeing you around here anymore. Sadly the War Room is no more and we are now retired but there were some memories and fellowship shared in there.
granger
Founder
I meant the idea truly grew from a project in the summer of 1987 when we were going to apply ELO to MLB data. Again only scores were available. We started the data collection in the library (talk about painful) and gf at the time had a Compaq 'laptop' (her dad was on the board because those things were expensive) where the front panel came off to be a screen (amber). Heavy as can be.
After two full days in the library, we gave up. You just couldn't get all of the scores. Some days the late scores would make the next day, sometimes not. It was a royal pain so we gave up.
Roll Hounds, Roll.
Maury P
Active member
I kind of skimmed through this thread so forgive me if my question has already been answered. How does the Toy calculate a forfeit? What is put in for both teams involved? Didn't really think of it until after this past Friday.
Mike
Administrator
I believe it's programmed to use the original spread if the forfeiting team was the underdog, and the spread was less than 45. Otherwise, it uses 45-0. I'd have to look at it to be sure but, if memory serves, that's how it works.
Maury P
Active member
Thanks for the info!
Didn't you tell us once many many moons ago that the inspiration for applying the ELO to MLB data was to use the results to bet on games and get rich quick?
granger
Founder
100%
Justlookin' I tried to send you a private message but after writing it when I tried to send it I got a notice I couldn't start a conversation with you.
So, I guess I'll go public with part of it here. I've nothing to hide that I can remember. Granger mentioned in passing several years ago hanging with the guys at the block plant. I never knew who those guys were but thought I might have met you in passing as we lived there from 2001-09 when we moved to Stephenville where I worked at Tarleton. My wife was raised there part of her elementary and jr hi years (JH cheerleader, football sweetheart 78) until they moved to the metromess after her dad passed. Their roots in Strawn go back to the 1800's after the Civil War. My brother-in-law was Donald (Dondi) Roberson who worked at the block plant the last several years of his life until he passed in I think 2010. Where was the war room?
Best regards,
Ron Bulman
justlookin'
Member
Without revealing too much, I'll admit to being one of two block plant employees Granger alludes to in his farewell article after the sale of the site to Mike. I worked there in the mid to late '70's and then left but returned in the mid '80's and was working there when the new ownership group came along in '91. I knew Donald and know a bunch of Roberson's. Attended Tommy Jack as well as Larry's funerals. The "War Room" was our conference room/kitchen at the plant. We had a map of Texas complete with stick pins in the locations of each and every sixman school in the state hanging on the south wall of the room. It later got moved to the east wall. Granger was impressed at our dedication to the game. We had also on that map the weekly rankings (done by the Lubbock AJ at the time) as well as our own sheet showing which one we agreed or disagreed with. we had every issue of Texas Football in the closet dating back to '63. I had all of them at one time but the '60 through '62 issues came up missing. Granger was there several times, always for a high-profile game. We enjoyed each others company and it didn't hurt that we were all UT fans. I'm not a Strawn native but I've lived here for over 40 years. Sorry about the PM thing , I didn't turn it off but it would appear that I never turned it on either.
Texlonghorn75
Active member
So were Tommy Levi Roberson and Donald Roberson brothers?
justlookin'
Member
I don't know.
No. Donald Ray (Dondi) has one brother. David. Who played end on Nelson Campbells 1979 first ever Strawn state semi team (lost to Milford by 20 (?)). Dondi, bless his soul wasn't a big football fan. Played 1-2 years. Never heard of Tommy Levi.
Tigerfan85
Active member
Ok so yes Westbrook lost but by no means does May need to leap frog to 1st. So just dial in whatever you need to to make sure Abbot, Rankin, Jonesboro or Garden City moves to that 1st spot JUST NOT MAY !! Please!!!!I mean, I was just joking but I'm not above taking a bribe!
I mean, I was just joking but I'm not above taking a bribe!
The #1 ranking curse!
