Attendance Analysis (Pt. 1): Introducing Att+

by Steve Slowinski on Jan 17, 2010 7:00 AM EST

When attempting to analyze the Rays' attendance issues, I've always found myself with more questions than answers. We all know that the Rays have had low attendance figures ever since their inception, but in my opinion, it's not possible to draw meaningful conclusions from the raw data. Looking at raw attendance figures is about as evaluative as looking at RBI totals. Do the numbers tell you what happened in the past? Sure they do. However, do they answer the question "Why?" Do they provide context? Do they evaluate a player/team's talent level? Eh, not so much.

The problem with raw attendance figures is that there are lots of variables that influence them. Does the team have a recent history of winning or losing? Has the team been established in the area for an extended period of time? How large is the metropolitan area? How high is the unemployment rate in the area? What are the demographics of the area? All of these variables can influence attendance figures, but are not accounted for when looking at the raw data.

In order to engage in a meaningful analysis of attendance issues, we need to bear these variables in mind and attempt to account for them to some extent. That's easier said than done, though, and it can make for lengthy articles full of confusing charts, baffling graphs, and dry text that only the excessively nerdy can stand to drudge through. To make it somewhat more bearable, for the next couple of weeks I plan to tackle these variables one at a time, slowly expanding our knowledge of attendance figures and hopefully arriving at some conclusions about the Rays' attendance. At the moment, I have no idea what those conclusions are going to be or if we'll be able to conclude much at all, so it'll be our own little adventure.

Since this topic is one that little research has been done on yet, please feel free to offer suggestions and ideas as this series goes along. I don't claim to be an expert on this subject - not even close - and anything you have to add is greatly appreciated. My final goal is simply a better understanding of the Rays' attendance situation and the more brains at work on the subject the better.

***

To begin the series, I want to introduce a new statistic that will be vital to our analysis: Att+. Much like the statistic OPS+, Att+ (or "Attendance Plus") is a measure of how a team's attendance figure compared to the league average for that season. One-hundred is league average; scores lower than 100 are below league average and scores higher are above average. For example, last year the Rays' average attendance per game was 21% lower than the league average, giving them a 79 Att+. During the same year, the Yankees' attendance per game was 57% higher than league average, good for a 157 Att+.

While Att+ doesn't control for many of the variables that we're concerned with, it does one very important thing: it allows us to compare team attendances across baseball history. If we're going to draw conclusions about attendance data, we need to be able to compare the Rays with other teams in history and we need to do so accurately. For example, compare the raw attendance figures (Att/Game) for the Rays' first three seasons (1998-2000) with the Mets' first five seasons (1962-1965):

Season	Mets	Rays
1	11,532	30,942
2	13,335	19,294
3	21,129	18,121

Looks like the Rays started off a lot stronger then the Mets, right? The Rays did drop off while the Mets increased over time, but the Rays outdrew the Mets handily those first two years. Now, though, let's look at those same seasons but using Att+:

Season	Mets	Rays
1	91	110
2	110	69
3	168	64

Holy cow! Both the Rays and Mets drew around league average their first seasons, but the Mets dominated the Rays over the next two years. Their league average attendance was around 12.5K per game while the 1998 league average was 28K per game, which makes quite a difference when comparing teams from those two decades.

Like the Mets, the Rays are an expansion team and face the problem of developing their fan base from the ground up. How have expansion teams fared during their inaugural years? Do all expansion teams struggle initially? Are there any teams that offer interesting comparisons to the Rays' current situation? Later on today I'll post the second part of this series, which will focus on these questions. Until then, here's a comparison to think about:

First 12 Seasons of Existence (Att+):

Season	Marlins	Rays
1	126	110
2	109	69
3	99	64
4	84	55
5	108	48
6	76	48
7	61	55
8	53	47
9	54	56
10	37	54
11	60	71
12	74	79

Tweet Comment 11 comments | 0 recs |

Do you like this story?

Comments

Attendance

I think this will be a tough subject to extract quantifiable reasons why attendance is low. I personally think, and have said this before, that although unemployment is important it may be equally important on what sectors are unemployed. It is my speculation that season tickets are often bought in the Tampa Bay area by companies tied to the housing industry. Plumbing companies, A/C companies etc., even furniture companies. I doubt the figures are available, but it would be interesting to know who was buying the season tickets, how many stopped buying and what industry they were representing. I doubt area unemployment figures counting accountants would have the same impact on baseball attendance as counting unemployment of concrete companies.

Another interesting aspect of attendance would be how many of the seats had been held by season ticket holders, and how many were not. My guess would be a loss of season ticket holders and proportionally more individual sales, as both individuals and companies decide to buy tickets to fewer games.

Good luck in your quest. I have no idea how you can figure this one out for sure.

Gary Williams for President!
Put Rose in the Hall of Fame

by terp12 on Jan 17, 2010 10:10 AM EST reply actions

I don't really intend to come up with a giant conclusion at the end

The idea is more to explore lots of historical comparisons and see how the Rays fit into the overall picture. My thought is that by looking a bit more broadly, we can gain a bit more of an appreciation for how fan bases develop and how attendance typically tracks in teams. Untangling all the tiny variables is virtually impossible on a small scale, but I’m hoping it proves a bit easier when you take a larger sample.

"I never threw an illegal pitch. The trouble is, once in a while I toss one that ain't never been seen by this generation." - Satchel Paige

by Steve Slowinski on Jan 17, 2010 10:34 AM EST up reply actions

Interesting stuff, Steve

I like the idea of ATT+, but it’s hard for me to just look at the data and extract anything meaningful. Could you run a correlation in your comparisons? It might make it easier to see which teams compare favorably and which ones don’t. For example, running the Rays/Marlins from above I get r=.53. That in itself doesn’t tell us a whole lot, but once you’ve got enough other teams you can rank the correlations to see which ones are most, and just importantly least, correlated to the Rays.

I'm a writer.

by Andy Hellicksonstine on Jan 17, 2010 11:41 AM EST up reply actions

That's coming up.

I was planning on doing comparisons with other teams probably next weekend or sometime this week.

"I never threw an illegal pitch. The trouble is, once in a while I toss one that ain't never been seen by this generation." - Satchel Paige

by Steve Slowinski on Jan 17, 2010 8:02 PM EST up reply actions

Ok

In that case, I think you will have a problem drawing any meaningful conclusions at all. Too many variables that could be of significant importance. How does one account for the stadium differences, access to games, winning, tradition, unemployment, TV exposure, etc.? I also agree with the post below from raysbran if you are going through the exercise. I don’t think Tampa today, compares well with the NY Mets in the 1960’s.

Gary Williams for President!
Put Rose in the Hall of Fame

by terp12 on Jan 17, 2010 11:50 AM EST up reply actions

Winning and tradition are easy to account for.

As for the others, I think some of them are partly accounted for by Att+. Yes, new stadiums, unemployment, and TV have influenced attendance in teams, but they’ve also been variables that have affected the entire league at the same time. Not exactly, but when you look at events through a historical lens, compare to league average for each year, and use a large enough sample, small variables like those should fall away.

"I never threw an illegal pitch. The trouble is, once in a while I toss one that ain't never been seen by this generation." - Satchel Paige

by Steve Slowinski on Jan 17, 2010 8:06 PM EST up reply actions

Not sure the Mets are a good team to compare this too

The Giants and Dodgers left NY in 1957, and the area was starved for National League baseball again. The NY metropolitan area is also very territorial. People in Queens and Brooklyn, at least back then, did not want to root for an AL team, especially one in the Bronx. They embraced the Mets, and especially with the Yankees being in the World Series in 1960 thru 1964, and with the Mets being so horrid, the lovable losers became the underdogs and gathered even more fans. Also Shea Stadium opened in 1964 which probably accounts for that spike in attendance in year 3.

Maybe you should compare the Rays to the Diamondbacks, Rockies or Marlins, teams that were put in geographical areas that never had teams before, unlike NYC. Just my $.02.

by raysnbran on Jan 17, 2010 11:24 AM EST reply actions

First 12 seasons, Ari, Col, TB

Season ARI COL TB
12 87 96 79
11 96 104 71
10 88 121 54
9 83 131 56
8 82 139 47
7 104 149 55
6 124 161 48
5 141 172 48
4 113 181 55
3 124 188 64
2 129 184 69
1 153 179 110

Correlations look like this:
TB/FLA: r=.53
TB/ARI: r=.39
TB/COL: r=.03

I'm a writer.

by Andy Hellicksonstine on Jan 17, 2010 12:13 PM EST up reply actions

Colorado was friggin' ridiculous.

An outlier to a huge degree, much like the Mets were in the 1960s. I had no idea they were embraced so readily early on…then again, they also draw fans from a large area, so that helps.

"I never threw an illegal pitch. The trouble is, once in a while I toss one that ain't never been seen by this generation." - Satchel Paige

by Steve Slowinski on Jan 17, 2010 8:27 PM EST up reply actions

See the second post.

Yes, I agree that the Mets aren’t a good comparison. It was just an easy example to see the difference between raw attendance and Att+…that’s all.

"I never threw an illegal pitch. The trouble is, once in a while I toss one that ain't never been seen by this generation." - Satchel Paige

by Steve Slowinski on Jan 17, 2010 8:03 PM EST up reply actions

I wonder if median per game attendance isn't a better baseline for ATT+

Averages can be skewed by big outliers, and there are a few of these that likely skew the data upward – the Yankees of the last few years for example, with attendance perhaps double the league average, a figure that can move the entire average up almost a 1000 per game. or 3-4 %. Compared to the low outlier, the Fish, who might come in only about 12k below median, or -400 per game per team. And I suspect there may be more teams varying higher than lower.

For example, in 2008, median attendance would be around 30,930 (midpoint between teams per game of 15th and 16th ranked). Only 1 club, the Marlins, were more than 11K below this, at 16,700ish. But the top 2 teams both exceeded the median by over 20K, and 5 were up more than 11K per. So the average might not be average.

Fans don’t move from 1 stadium to another by and large, so attendance largely happens in discrete packets within markets. Median just seems a more appropriate number than average intuitively, but it may not make much difference in the grand scheme of things mathematically.

That said, your approach is a good one, suggesting to me that there may not be quite as much upside in Rays attendance possibilities than I might have thought.

by nyyfaninlaaland on Jan 20, 2010 3:30 AM EST reply actions