Attendance Misconceptions

[Quazzum69]

There's a lot of conjecture around here about football attendance and its decline; I thought I'd clear some things up before another dozen threads start.

1. Basketball attendance and football attendance are inversely related

NO, quite the opposite. There's really not much correlation between average basketball attendance and average football attendance. There's actually a slight positive correlation between the two. This may suggest people's overall ability and willingness to attend SU athletics in general change, but not for specific sports (must choose between football or basketball). Maybe the athletic department as a whole has "good" periods and "bad" periods.

[A

 

[Quazzum69]

2. Time of kickoff has a significant effect on attendance (early games don't draw well)


NO. Since 2001, there is no difference in attendance between games at different times. The average attendance for 12:00 pm games is 38,780 (32 games), between 1 pm and 2 pm averaged 40,880 (11 games), between 3:30 and 4:30 averaged 37,600 (13 games) and between 6:30 and 8 averaged 38,500 (19 games). However, there hasn't been any 1 - 2 pm games since 2006, which excludes the two worst years (2007, 2008). What's really important is that there is no difference in attendance between early noon games and 7 or 8 pm games.

 

[texascpa]

How dare you provide charts and numbers to support your thesis. It's outrageous!

 

[dollarbill44]

What's going on here?:noidea:

Final attendance figures don't show the amount of people in their seats at game time, which is definitely impacted by start times.

 

[stuckinbig11]

R squared isn't great, but the trend is there.

 

[MikeSU02]

Quazzum69 Thanks for posting these. Keep it coming.

 

[Quazzum69]

3. Streaks are important (more people attend when team is on a winning streak than losing streak)

There's not much, if any, difference when the team has consecutive wins or losses prior to the game (2001 to 2012). Exceptionally long streaks (four losses, seven wins, etc.) are common in really good years or really bad years and since winning percentage is a pretty good indicator of attendance, the streak is probably not the cause of changes in attendance.

 

[BlazeOrange]

What's going on here?:noidea:

Final attendance figures don't show the amount of people in their seats at game time, which is definitely impacted by start times.

Not even just at game time if paid attendance vs. gate attendance was used (OP didn't specify, but I don't know if reliable turnstile data is available - definitely not as easy to find as the published paid numbers).

I'd also like to see this graph broken into three charts or color code the datapoints with the following categories:

1. Conference Games
2. Non-conference vs. BCS opponent
3. Non-conference vs. non-BCS opponent

Many of the games that SU has control over and can put in more favorable time slots are against opponents with less drawing power, leading to an apples/oranges comparison.

 

[upperdeck]

one thing we never really know is season tickets vs walkups vs presale. assuming season tickets were mostly sold before game 1 thats roughly 3k more for tulane vs wagner. how many were sold before the 2 losses ?

 

[Crusty]

Well, if you are going to use facts you have to be banned.

 

[NineOneSeven]

Fat finger trend.

 

[orangeinjersey]

How dare you provide charts and numbers to support your thesis. It's outrageous!

Theres no room for the truth on the internet

 

[Legendof44]

Is there room for how the economy in CNY is doing on that graph?

 

[PeteCalvin]

Theres no room for the truth on the internet

Where have you been? Everything on the Internet IS true. Sheesh.

 

[cusefan88]

Only thing is the attendance listed in a box score is tickets distributed but we all know "butts in seats" is always lower. For example, Tulane attn is listed as 36,128 however I heard from a good source actual fans in was around 25-26K.

Love the data but what SU Athletics lists as attendance is recently quite far from reality.

 

[JoeSU]

Nice analysis.
Box and whisker plots would have been better.
I kid (sort of).

 

[xc84]

Only thing is the attendance listed in a box score is tickets distributed but we all know "butts in seats" is always lower. For example, Tulane attn is listed as 36,128 however I heard from a good source actual fans in was around 25-26K.

Love the data but what SU Athletics lists as attendance is recently quite far from reality.

It is reality. It publishes paid attendance. People clamor for "sellouts" and do you know what that means? It means all tickets were sold. I laugh that so many people think they are the arbiters of "real" attendance as if they have some sort of expertise in estimating how many people are actually in the entire dome.

 

[cusefan88]

It is reality. It publishes paid attendance. People clamor for "sellouts" and do you know what that means? It means all tickets were sold. I laugh that so many people think they are the arbiters of "real" attendance as if they have some sort of expertise in estimating how many people are actually in the entire dome.

Well I looked at the info sheet Dome staff got before the Tulane game and it said they estimated attn between 25-26K, not the 36K they announced.

I care about the amount of people who actually go, tickets distributed is meaningless if people don't attend. It's a running joke about the differences in announced attn versus people in the seats.

How about Quazzum makes a new chart with attn from 2004-now. The glory years skew the reality of SU's current situation.

 

[orange79]

Well, the bar for opinions just got set a bit higher

 

[SWC75]

Maybe there's about the same number of SU football and basketball fans but a small football crowd is a big basketball crowd.

If you look at the reported crowds for the games I'm covering on my "Bold Brave Men of Archbold" series, (27,000 for the 1956 SU-Maryland game when the Terps were a top national power and SU and coming team), modern numbers don't seem so bad.

 

[Timonen]

Great info. Here's an angle that comes to mind, though...

Since the largest share of attendees are season ticket holders who will be there (on paper, at least) regardless of time, opponent, streak, etc., it's logical that none of the variables introduce show themselves as incredibly impactful. What if we looked at the deviation or difference from that ticket holder base? That's the variance that's making the decision based on those variables. Since lower n's, the diffs are more likely to show as significant.

Just a thought!

 

[xc84]

Well I looked at the info sheet Dome staff got before the Tulane game and it said they estimated attn between 25-26K, not the 36K they announced.

I care about the amount of people who actually go, tickets distributed is meaningless if people don't attend. It's a running joke about the differences in announced attn versus people in the seats.

How about Quazzum makes a new chart with attn from 2004-now. The glory years skew the reality of SU's current situation.

Paid attendance is not a joke at all. It's what every school announces. It is not "meaningless". If you sell something, do you not care how many were sold? That's meaningless? SU sells seats. They care more about how many people buy those seats than how many people actually use them. Granted, they care about both but if "fannies in seats" were the ultimate metric of success, the solution is simple...give the seats away for free (but apparently not to the students).

 

[xc84]

How about Quazzum makes a new chart with attn from 2004-now. The glory years skew the reality of SU's current situation.

I think that's a great idea. It would tell use the correlation between wins and attendance (and there is probably a year or 2 lag there).

 

[OrangeAl]

Is there room for how the economy in CNY is doing on that graph?

Is there some index to represent the local economy, like a GDP for central NY?
How about the total number of industrial jobs or something like that?

 

[reedny]

There's a lot of conjecture around here about football attendance and its decline; I thought I'd clear some things up before another dozen threads start.

1. Basketball attendance and football attendance are inversely related

NO, quite the opposite. There's really not much correlation between average basketball attendance and average football attendance. There's actually a slight positive correlation between the two. This may suggest people's overall ability and willingness to attend SU athletics in general change, but not for specific sports (must choose between football or basketball). Maybe the athletic department as a whole has "good" periods and "bad" periods.

[AView attachment 5170

Just wondering, in Figure 1, why didn't you plot FB/BB attendance together along the Y axis, and then do a chronology along the bottom (X) axis (similar to the way you compared data in Figure 2)? I can understand why you set it (figure 1) up this way -- its a simple way to plot the two attendance levels. But without time references for the data points, you can't compare BB/FB attendance season by season to determine correlation. Graph 3 is pretty cool.