End of an Era: Pablo
Jul 16, 2024 15:02:25 GMT -5
HawaiiVB, Cubicle No More ..., and 58 more like this
Post by The Bofa on the Sofa on Jul 16, 2024 15:02:25 GMT -5
All
After something like 20 years, it is time for me to hang up the Pablo shoes. It's been a wonderful experience, but I am not going to be able to continue this year.
I am really proud of what I have developed with Pablo over the years. It started as a simple model based on a concept that there was a relation between the outcomes of matches and how good the teams were. Originally, it was highly empirical. I can remember my curiosity when I saw that a "lop-sided 5 set match" had the same predictive power as an "average 4 set match" or a "close 3 set match." Of course, nothing will match the day when I was trying to simulate the NCAA tournament and needed to turn Pablo ratings into point probabilities, and seeing that straight line! All of a sudden, there was a foundation for the model, and from there, it took off (and, in fact, got to be a lot easier).
Over the years, Pablo has shown up in a lot of places.
1) I don't remember what year it was, but at one point I presented it to the American Statistical Society.
2) Oh, the effort to get the NCAA to replace RPI with Pablo. Working with the AVCA and coaches, we did everything we could. I spent probably a hundred hours working on that project. All for naught, unfortunately.
3) Pablo has shown up in other places. At one point, we were including JC schools. For many years, I did Pablo rankings for NCAA Men and for the NCAA Women's Beach (that was a lot of fun, because of the nature of the competition). At one point, I even talked to Karch about creating an individual rating for beach players (kind of like a chess rating). I couldn't figure out how to make it work, though.
4) In terms of the model, I have done a ton of analysis of results, and everything I do shows the model is robust. I know now that it has to be, given the mathematical foundation, but I've been able to figure out how to assess the quality at any point, and can see when (during the season) it works better or is worse. I have a database of more than 100K "matched pairs" where two teams play each other twice in the same season, so see how the outcome of the first informs the second.
5) My work on Pablo has gotten me to look at other sports to see how well I can do there. For a while, I was looking at women's tennis, thinking that it should be similar. Amazingly, it doesn't work nearly as well as it does for volleyball. I interpret it to the fact that tennis is an individual sport whereas volleyball is a team sport, such that the standard of deviation of play in tennis ends up being a lot larger (ultimately, that explains the outcome). In contrast, because volleyball is a multiplayer sport, if one person has a bad match, it can often be offset by another player having a good match. I've also looked into other sports (NFL, NBA), but the model is not quite the same.
6) Perhaps my favorite part is come tournament time, when Pablo Analyzes the Brackets. I love the idea that we can provide an objective foundation for those common questions, like "who has the easiest path to the final 4," etc. And it hasn't even always been Penn St!!!
For sure, working on Pablo has changed the way I view outcome and performance, and the entire nature of competition.
But after doing this for more than 20 years, I will be giving it up. Other responsibilities in life make it so I just can't commit to doing it every week any more.
I need to thank those people who have helped my thinking over the years and made this possible. mikegarrison and hebrooks87 always provided insightful comments, and Ruffda always made me think differently. Of course, none of this would have been possible without the help of RichKern, who not only gave Pablo a voice, but provided the data. Oh those early years when volunteers like Amy Peacock built the D2, D3 and NAIA databases! Some of you may remember the days before Pablo was on RichKern and was posting my first Pablo rankings right here on VT! (and that's why they are Pablo rankings, because that was my screen name at the time (before it was Agent Buchwald, which then turned into the Bofa)). But thanks to Rich, not only for giving me a platform, but for helping to lift NCAA women's volleyball for everyone. Thanks also to all the coaches I've met and talked with over the years. Introducing myself as Pablo always gave me an opening, and I've loved talking with them. Coaches love to talk about their teams, and I always enjoyed listening. And finally, thanks to all the players over the years who've made it fun to watch, especially those who have autographed my volleyballs for me. In the end, volleyball is still about the players on the court and what they do. The reason I got into doing Pablo, and the reason I continue, is because I am a volleyball fan. That will never change.
While I am giving it up, the good news is that I am in contact with a potential successor who has offered to carry the ball. I will do what I can to help bring that person up to speed on what I do - I'm sure there is a better way, so hopefully they can even make it easier! Hopefully, you will not even notice a difference, and there will be Pablo rankings ready for the first week. If not, please be patient, we'll get there. While it may not be me with the effort, the goal is to continue on. Wish us luck.
I don't hang out much here these days, but I won't be going away completely and you may see me around. I just won't be spending 6 hours every weekend doing Pablo calculations.
Best Wishes
The Poster Formally Known as Pablo
After something like 20 years, it is time for me to hang up the Pablo shoes. It's been a wonderful experience, but I am not going to be able to continue this year.
I am really proud of what I have developed with Pablo over the years. It started as a simple model based on a concept that there was a relation between the outcomes of matches and how good the teams were. Originally, it was highly empirical. I can remember my curiosity when I saw that a "lop-sided 5 set match" had the same predictive power as an "average 4 set match" or a "close 3 set match." Of course, nothing will match the day when I was trying to simulate the NCAA tournament and needed to turn Pablo ratings into point probabilities, and seeing that straight line! All of a sudden, there was a foundation for the model, and from there, it took off (and, in fact, got to be a lot easier).
Over the years, Pablo has shown up in a lot of places.
1) I don't remember what year it was, but at one point I presented it to the American Statistical Society.
2) Oh, the effort to get the NCAA to replace RPI with Pablo. Working with the AVCA and coaches, we did everything we could. I spent probably a hundred hours working on that project. All for naught, unfortunately.
3) Pablo has shown up in other places. At one point, we were including JC schools. For many years, I did Pablo rankings for NCAA Men and for the NCAA Women's Beach (that was a lot of fun, because of the nature of the competition). At one point, I even talked to Karch about creating an individual rating for beach players (kind of like a chess rating). I couldn't figure out how to make it work, though.
4) In terms of the model, I have done a ton of analysis of results, and everything I do shows the model is robust. I know now that it has to be, given the mathematical foundation, but I've been able to figure out how to assess the quality at any point, and can see when (during the season) it works better or is worse. I have a database of more than 100K "matched pairs" where two teams play each other twice in the same season, so see how the outcome of the first informs the second.
5) My work on Pablo has gotten me to look at other sports to see how well I can do there. For a while, I was looking at women's tennis, thinking that it should be similar. Amazingly, it doesn't work nearly as well as it does for volleyball. I interpret it to the fact that tennis is an individual sport whereas volleyball is a team sport, such that the standard of deviation of play in tennis ends up being a lot larger (ultimately, that explains the outcome). In contrast, because volleyball is a multiplayer sport, if one person has a bad match, it can often be offset by another player having a good match. I've also looked into other sports (NFL, NBA), but the model is not quite the same.
6) Perhaps my favorite part is come tournament time, when Pablo Analyzes the Brackets. I love the idea that we can provide an objective foundation for those common questions, like "who has the easiest path to the final 4," etc. And it hasn't even always been Penn St!!!
For sure, working on Pablo has changed the way I view outcome and performance, and the entire nature of competition.
But after doing this for more than 20 years, I will be giving it up. Other responsibilities in life make it so I just can't commit to doing it every week any more.
I need to thank those people who have helped my thinking over the years and made this possible. mikegarrison and hebrooks87 always provided insightful comments, and Ruffda always made me think differently. Of course, none of this would have been possible without the help of RichKern, who not only gave Pablo a voice, but provided the data. Oh those early years when volunteers like Amy Peacock built the D2, D3 and NAIA databases! Some of you may remember the days before Pablo was on RichKern and was posting my first Pablo rankings right here on VT! (and that's why they are Pablo rankings, because that was my screen name at the time (before it was Agent Buchwald, which then turned into the Bofa)). But thanks to Rich, not only for giving me a platform, but for helping to lift NCAA women's volleyball for everyone. Thanks also to all the coaches I've met and talked with over the years. Introducing myself as Pablo always gave me an opening, and I've loved talking with them. Coaches love to talk about their teams, and I always enjoyed listening. And finally, thanks to all the players over the years who've made it fun to watch, especially those who have autographed my volleyballs for me. In the end, volleyball is still about the players on the court and what they do. The reason I got into doing Pablo, and the reason I continue, is because I am a volleyball fan. That will never change.
While I am giving it up, the good news is that I am in contact with a potential successor who has offered to carry the ball. I will do what I can to help bring that person up to speed on what I do - I'm sure there is a better way, so hopefully they can even make it easier! Hopefully, you will not even notice a difference, and there will be Pablo rankings ready for the first week. If not, please be patient, we'll get there. While it may not be me with the effort, the goal is to continue on. Wish us luck.
I don't hang out much here these days, but I won't be going away completely and you may see me around. I just won't be spending 6 hours every weekend doing Pablo calculations.
Best Wishes
The Poster Formally Known as Pablo