Responding to polo86:
The Polo Coach website (polocoach.com/cifss/ratingsreport.php) does not provide any information about the methodology it uses to create its polls or where it obtains the scores and other data it uses for its polls. Therefore, one is left to speculate about its methodology and data. Although I haven’t studied Polo Coach’s September 24 boys’ high school water polo poll (Poll) in great detail, it leaves a lot to be desired. Reasonable people can disagree about whether Sacred Heart should be ranked in the top 5 nationally and whether Greenwich, San Clemente, and La Jolla should be ranked in the top 20. However, no serious observer of high school water polo would rank Buena and Arcadia ahead of Sacred Heart or rank De La Salle 24th, as the Poll does.
Computer programs are only as good as the data they are provided or, in the case of sophisticated programs, can find on their own. As some in the computer programming world used to say: GIGO (garbage in, garbage out). I’m not sure where the Polo Coach obtains its data but the records shown in the Poll for many of the teams are inaccurate, some wildly so. To cite just three examples, the Poll ranks De La Salle 24th with a record of 1-2. De La Salle is actually 6-2. Second, the Poll ranks La Jolla 6th with a record of 8-1. La Jolla is actually 15-3 but it hasn’t won a game against any team in my top 20 list and lost games to Rio Americano (ranked 155th in the Poll) and Marin County’s Redwood (not ranked in the top 322 in the Poll). Third, the Poll ranks Oaks Christian 8th with a record of 6-3 and a 16-13 loss to Harvard-Westlake on September 14. Oaks Christian is actually 7-2 and beat Harvard-Westlake 16-13 on September 14. I thought the Poll might be pulling its data from the CIF Southern Section’s website but that website correctly shows that Oaks Christian beat Harvard-Westlake on September 14. The MaxPreps website and the Poll show the same incorrect score for the September 14 game but the MaxPreps website correctly shows that De La Salle’s record is 6-2, not 1-2 as the Poll shows.
The major sportsbooks used to allow (and may still allow) a small number of professional gamblers to bet on sporting events before the general public was allowed to do so. The purpose of this exercise was to truth test the sportsbooks’ initial point spreads and odds. If all (or a significant percentage) of the professional gamblers bet the same way, the sportsbooks would adjust their points spreads and odds to reflect the feedback. One doesn’t have to run a sportsbook to benefit from truth testing a product before it’s released to the public. Moreover, one doesn’t have to know much about high school water polo to be suspicious of a poll that ranks a school as the 6th best team in the country even though it has lost to the team the same poll has ranked 135th and to a team the poll doesn’t have ranked in the top 322. My unsolicited advice to Polo Coach is to ask Joel Francisco or someone else knowledgeable about Southern Section water polo to review its polls before they are released to the public.
Judging from the title on its website, “CIF-SS Power Ratings Tool,” it appears the Poll is designed to be used or considered by the CIF’s Southern Section. If this is true, the poll would be more credible if it didn’t attempt to rank the non-Southern Section schools.
To be fair to Polo Coach, the MaxPreps polls aren’t very good either. For example, MaxPreps ranks Oaks Christian 33rd in California. Oaks Christian is easily a top 10 team. MaxPreps also ranks Oaks Christian 41st nationally. That’s ludicrous. There are probably no more than 3 non-California schools that would be ranked in California’s top 30 or 40: Brunswick (the bottom half of the top 10), Greenwich (somewhere in the 13 to 20 range), and Punahou (somewhere in the 30s or early 40s).
Water polo is not as amenable to statistical analysis as many other sports. Part of this is because record keeping in water polo is often haphazard or inaccurate. It’s difficult to obtain accurate schedules and scores from some schools, let alone the kind of data one would need to rank teams and players as is done in sports such as baseball, football, and basketball. Nevertheless, I believe it would be helpful for a small group of coaches and knowledgeable observers to attempt to write a computer program that would rank the schools more accurately than the existing programs.
