More photos » by Lynne Sladky - AP

Thank goodness we're not the Marlins... (AP Photo/Lynne Sladky)

Browse more photos »

For those of you that have been following my series of posts on attendance issues, I've run some more numbers on how the Rays have historically drawn versus the other teams in the AL East.  A couple quick points to remember about these numbers:

1. These numbers are measured in percent-above-average, so for example if we drew +6% against a team for a year, this means when we faced them, we drew 6% better than our average attendance per game for that year.  I determined these percentages for every year and then created a weighted average of these years to determine how we're drawn against each of these teams going back to 1998. 
2. These numbers have almost no predictive value.  Attendance against certain teams fluctuated wildly from year to year, mostly because there are a variety of other variables that act upon attendance (like day of the week, holidays, weather, give-aways, etc) and we're working with small sample sizes within each year (for example, we only faced the Red Sox nine times at home last year).
3. With this in mind, though, they do give us a good idea of what to expect generally when facing each team.  If a team has a -10% figure, we should not expect to draw well against this team, even if we are facing them on a weekend.  And if we happen to be facing them at home on the weekday, we should anticipate very low attendance.

I highly doubt that those numbers come as any surprise to anyone.  Oh wow, we draw well against the Yankees and Red Sox?  No way!  I was surprised, though, by how similar the numbers for the Orioles and Blue Jays were (aren't the Blue Jays always more competitive than the Os?), and that we draw significantly better against the Yankees than Red Sox.  My impression has always been that there are a ridiculous amount of Red Sox fans in the Tampa area, but it appears that there are just as many Yankee fans there as well.  Maybe the Yankee fans are simply less obnoxious...

When looking at these numbers, I couldn't help but wonder if maybe they were influenced slightly by scheduling.  For instance, are the Rays more likely to be scheduled against the Yankees for a weekend series, which would therefore boost attendance even further against them?  When looking at it, though, I found quite the opposite.  What follows is a chart of the number of games since 2004 against these opponents that fell either between Monday - Thursday (low attendance days) or Friday - Sunday (peak attendance days).

Wow, this just makes our attendance against the Yankees and Red Sox even more impressive.  Not only are we drawing significantly higher than our yearly averages against them, but we've drawn that amount on a preponderance of Monday - Thursday games.  Is this scheduling an intentional?  Possibly not since the Rays' FO doesn't have any say in how the yearly schedules are planned, but regardless of if it is or not, it works out in favor of the Rays.  With their schedules planned this way, the Rays are able to draw fans out to the park during the weekdays, while counting on attendance on the weekends to still be high.  In short, they are optimizing how many people they are drawing to the park during a given season.

Also, with all this talk about the inter-relationship day of the week and opponent, I wanted to know which variable is more powerful when determining attendance.  To answer this question, I ran a linear regression analysis on SPSS to find the regression coefficients for each of these variables.   While I'm a bit rusty on SPSS and need to spend some more time with it, I found that when using the 2008 season as my sample, I got a .274 coefficient for Day of the Week and a .195 coefficient for Opponent.  In other words, these results suggest that day of the week accounts for 27.4% of the variability in attendance numbers and opponent accounts for a further 19.5%.  In total, that means with just those two variables, we're accounting for around 50% of the variability in attendance numbers, with day of the week being more important than the opponent faced.  Anyway, this is still in need of more fleshing out, but I believe this gives at least partial credence to the focus on opponent faced and day of the week because with only these two variables, we're accounting for most of the variability in attendance.

I don't claim to be a statistics master, so anyone with advice or comments on regressions and/or SPSS, please feel free to leave some tips.  I'd love to keep getting more specific and fleshing this analysis out even more.