Truthfully I have never anxiously awaited my copy of Bowler’s Journal until this month, when of course I had to wait for it to be forwarded since we just moved. I got a call from George Aboud (he of mystery hand problems) telling me that Mike Fagan had written on statistics and averages. How cool! Since I wrote on fixing the average system earlier (for the record before the Moneybowl issue came out), I was hotly anticipating what Mike had to say.
Bowling Center “Slope.” In the article Fagan mentions a bowling center “slope,” that is, an adjustment for each center based on the scoring pace for that particular house. So, for example, if a bowler averages 210 in one house and 214 in another, there should be a normalization factor that makes those two averages seem equal. He dismisses oil pattern as a non-factor and focuses solely on a house-to-house adjustment, and since most leagues bowl on easy house shots this makes sense.
The slope idea is perfect but I do see oil pattern as part of the equation for several reasons. First, any competent lane man can put out a tough or an easy shot, making the same house play like two entirely different centers. Secondly, we want to ENCOURAGE bowlers to request different, more challenging patterns so they can see how they fare on different conditions at the same center (and become better bowlers). Finally, many tournaments put out different patterns than the house shot and need a way to normalize averages and handicaps beyond letting sandbaggers win and then re-rating them. Oil pattern is THE ONE additional piece of data USBC has to track and the rest we can derive through the magic of computers and data analysis. The good news is that factoring in oil pattern isn’t hard – all we have to do is record the pattern in play for any sanctioned score bowled, even if it reads “house shot.”
Deriving “Pattern-to-Par” (PTP) multipliers. Once we have this data we can, as Mike points out, analyze scores for bowlers in that and other centers, with one big lynch pin being the same bowlers who bowl in multiple leagues in multiple centers. From here we derive our PTP (Pattern to Par) multipliers, which allow us to MULTIPLY an average to normalize that average to par, then DIVIDE to take the normalized average and predict what that same bowler will average at a different house and/or on a different pattern. Normalizing an average to par (i.e. multiplying it by the PTP) is the first step in comparing all bowlers to the same standard and achieves the unbiased comparison of players Fagan is looking for.
Let’s Get Started! How do we start? Easy. USBC has all the data – all we have to do is extract it and start analyzing. We need them to change their average databases to include oil pattern. For previous years, we can generally just say “house shot” for most leagues and get the starting point we are looking for. PTPs will evolve over the years and become more accurate as scores from more different patterns are recorded. The rest are computer algorithms that aren’t very hard to write and apply.
We Have The Technology. Len Nicholson told me I was crazy to try and implement this idea. Others have told me that “bowling isn’t like golf, you don’t have the same talent and you will never get this done.” Hogwash! National champion Mike Fagan is calling for change. I am a 25-year software and database guy. My fellow junior coach Mark Schipp is an applied math expert. We have many friends who are Information Technology experts and are passionate about bowing. I don’t speak for Mark, but I am willing to contribute time and expertise on how to do this, (read: fly me down to Arlington and we’ll brainstorm) both in analyzing prior years’ scores to arrive at baseline PTPs and how to start recording and analyzing the right data to refine those PTPs going forward. It wouldn’t cost a lot of money, and wouldn’t be that hard! The main obstacle is USBC acknowledging the need for such a system and asking for help from the bowling community to get it done. Mike Fagan thinks this is a good idea and so does every other bowler I’ve ever explained it to. LET’S GET STARTED!