Citizens Bank Park, Philadelphia (MLB #17 - Cheers for Philadelphia fans)

 

Last night, we went to a game in Philadelphia, with a nearly-packed house (41,882) on a Monday evening. Kudos to the fans for turning out in droves – the Phillies have been near the top of the majors in attendance for the last several years. The Phils have also been in the post-season for the last three. In other words, this looks like a model franchise in a lot of respects.

But for most of the first century of their existence, the Phillies were frequently the anti-George Washington, last in the majors in wins, last in attendance, and last in the hearts of even their own fans, who booed them mercilessly (more on that tomorrow). For example, from 1920 through 1942, they had fewer seasons where they won as many as they lost (one, a year where they won 78 and lost 76) than they had where they won 30% or less of their games (five), and their total attendance was 55% or less of the league average every year.

How do you get from point A (1942) to point B (today)? I can understand how you can build a winning team from time to time, but how do you go from irrelevance to bringing in the fans as consistently as Philadelphia does now? 

How do you break the cycle of losing in front of small crowds?  

This is part of my fascination with what drives attendance figures (a more numbers-oriented discussion is after the notes about tonight’s game). For the last couple of years I’ve been tinkering with the numbers, trying to understand that. Last year, I did an analysis of several years’ worth of attendance figures, with a spreadsheet that included attendance totals along with the factors I could think of that might influence them (stadium size and age, metropolitan population, winning percentage, last year’s winning percentage, last year’s success in the playoffs, success in the playoffs in the last 20 years, last World Series won, payroll, last year’s attendance, even the average age of the players), and for the five years where I looked, in every case, the factor that correlated most strongly with attendance was last year’s attendance. In one way that makes sense, but if you look at it from a business point of view, it’s a chicken-or-the-egg problem –  the best way to have big crowds is to have had big crowds last year. That would seem like the Phillies of 1942 (or even of 2000, when they won 40% of their games and had an attendance that was only 65% as high as the average in the league), would be doomed forever, but they clearly weren’t. And they had some high attendance years between 1972 and 2000. 

The last few days, I’ve been looking at the problem two other ways. First, I looked at what the change in attendance correlates with each year. Then, I looked at what the attendance of one team (first the Phillies, later the Rays) correlates with over many years, instead of looking at what the attendance of many teams correlates with over one year.

One hint about breaking the cycle was already in the data that I’d done. The factor that came in 2^nd every year in terms of correlation with attendance was the team’s payroll. I would have thought that it would be one of the ones related to on-field success, like winning percentage (this year’s or last’s) or last year’s postseason success or something like that, but it wasn’t. And payroll doesn’t correlate very well with performance in any single year, although it does to some extent. When you look at what changes in attendance correlate with, payroll is again a bigger factor than winning, though it isn’t terribly strong. But when you look at the Phillies since 1998 (when I could find payroll numbers), their attendance correlates with payroll in about the same way that it does for the majors as a whole in a single year, and when you look at their attendance over the years, it correlates with their winning percentage very strongly, more strongly than with payroll, far more strongly than winning correlates with payroll (though they do correlate), or than attendance correlates with winning for the majors as a whole in a single year.

I think what’s going on is that winning does matter for attendance, but within the constraints of a team’s market. The Tampa Bay Rays have never been above 17^th in the majors in attendance, and only once was the payroll above 14^th, but their attendance correlates with winning even more strongly. The Rays win by finding great young players they don’t have to pay much, the Phillies win by finding great old players that they pay a lot. For a team in a reasonably big market, like the Phils, I think that a winning season brings more fans, which gives the team more money to sign good players, which makes them more likely (though not guaranteed) to win more, and if they do, they get a bigger attendance, so they get more money, so they can sign more high-priced players, etc.

Attendance does correlate with the team winning on the field, in the sense that when the team wins more, more people come out, and the opposite when they lose more. But when looking across the 30 major league teams for a single season, that is obscured by the fact that different teams are at different places in the cycle. If the Phillies do worse this season than last, they’ll probably draw fewer fans, and if someone near the bottom of the attendance list does better, they’ll probably draw more, but the two teams could cross in wins far before they cross in attendance because going (or not going) to games has a strong habitual part to it.

But back to the big question: how did the Phillies go from a losing team with a low payroll drawing few fans to one of the most successful franchises, by all those measures, of the mid-2020s? One thing that can jumpstart attendance is a new stadium. The Phillies have gotten new stadiums in 1971 and 2004. Their attendance doubled relative to the league average the first time, and increased by more than 20% the second time. Along with that, they started building good teams that kept that attendance up. The two times the Phillies have won a World Series are the 10^th year at Veteran’s Stadium, and their fifth at Citizens Bank Park. But each time, they later had a streak of multiple years where attendance was well below league average and winning percentages stayed below 50%. The most recent streak like that ended abruptly in 2019. 

 The 2019 turnaround

Why did things change after 2018? The Phillies did the other thing that should lead to a jump in attendance – they spent a lot of money over the offseason. They signed three free agents to big long-term contracts, J.T. Realmuto, Zach Wheeler, and, most notably Bryce Harper, the most coveted player that winter. Spending big doesn’t always pay off (the Angels are proof of that), but in this case it worked, because those three are all still performing at a high level with the Phillies (and were crucial in last night’s game). So the team is winning, they’re signing more high-priced players, and the fans are coming.

One thing to remember, though, is the cyclical nature of things. The Phillies have been spending big, and winning, which means that they have also spending wisely, and perhaps with some good fortune. But unless they have the resources to make mistakes, but then spend even more to fix them, like the Yankees or Dodgers, they will probably again have bad stretches.

For now, though, it’s a good time to be a Phillies fan.

The game

It was an exciting game to watch, with the Phillies winning in extra innings, beating the Chicago Cubs 4-3 in 11.

As if to emphasize my point about the cyclical nature of success for most teams, this was a matchup of the teams with two of the four best records in the National League at the moment, with both looking likely to make the playoffs. But when I was growing up, they were two of the least successful teams in the majors. The Cubs famously went 107 years between winning World Series, from 1908 through 2015, the longest such streak in major league history. The Phillies longest streak of phutility is “only” fourth on the list, at 77 years, from the start of the interleague championship in 1903 until 1980. But the Phillies had been playing in the National League for 21 years before that, without winning a league championship (in a league with 8 to 12 teams), so it’s comparable. But now, they’re both good.

For my money, the star of the game was Zach Wheeler, one of those 2019 free agents. He got the 1000^th strikeout of his career as a Phillie, which puts him 10^th on the list among all the pitchers who’ve worn a Phildadelphia uniform in their 140+ years. He gave up a home run to the second batter, but only two other hits, both harmless singles, and a walk in the six innings he pitched.

On the other hand, the rest of the Phillies looked sluggish at best. In the first three innings, they had one runner picked off first, had another thrown out at second by an outfielder while trying to take an extra base (another Phillie got thrown out in the 8^th trying to do the same thing, the first time I’ve seen it happen twice in a game in a long, long time), and had committed an error in the field.

But Wheeler kept the Cubs in check, and the Phillies eventually scratched out two runs. A relief pitcher gave up the tying run in the 8^th inning, but in the 11^th, the Phils got two good bunts that both turned into hits (maybe the Cubs misplayed one of them), and two well-hit balls out of the reach of outfielders to win it, with Realmuto (another of those 2019 free agents) scoring the winning run.

It was a good night to be a Phillies fan.

The company we keep

On our quest, we’ve gone with old or new friends on more than a dozen occasions. Last night, it was a new friend, Eli Harvey. Eli just finished his first year of graduate school at Penn, so he lives in Philadelphia. Apparently he and I met briefly four years ago at a scientific conference, but we’d certainly not spent time together. However, Eli’s father, Ralph, ran the Antarctic Search for Meteorites (ANSMET) for many years, and I spent four years camping out in the interior of Antarctica with Ralph and/or his team, plus I’d known him for a dozen or more years before that. When I was talking to Ralph about this trip, he suggested that Eli would be interested in coming to a game with us. I’m not sure I would have been outgoing enough when I was in my early 20s to go to an event with one of my parents’ friends whom I hardly knew, but Ralph assured me that Eli was more adept socially than he (Ralph) or I, and was correct. It was a delight.    

Now, back to…

What do changes in attendance correlate with?

One approach I took was to look at what correlated with the change in attendance each of those years. Since things like stadium sizes and city populations don’t change much from year to year, that limited it to things like payroll and on-field success. Also, since the 2020 season was played without any fans in the stands, and 2021 was played with restrictions on the number of fans in attendance, variable from team to team, I really could only do the “change” analysis for three years from the data I had already tabulated, 2024, 2023, and 2019.

The table lists the “R²” value that Excel spit out in each case for the factors involved in change. By way of orientation, 0 means they’re not correlated at all, a positive number means that when one goes up, the other does too, and 1 means they basically fall on a straight line. By way of comparison, for 2019 through 2024, the R² for the correlations of attendance (not change in attendance) with last year’s attendance ranged from 0.70 to 0.92, the correlations with payroll from 0.56 to 0.75.

Year	Attendance change vs. Change in wins	Attendance change vs. Payroll change	Payroll change vs. Change in wins
2024	0.19	0.15	0.02
2023	0.13	0.20	0.35
2019	0.02	0.32	0.35

 

None of the “change” correlations are as high as the correlations between this year’s attendance and last year’s attendance or even as high as the correlations between this year’s attendance and this year’s payroll. But for the three years I looked at, either a bigger payroll or winning more games correlated with a bigger attendance, although in one case, 2019, the correlation with wins was miniscule.

Incidentally, a bigger payroll correlated with winning more games, although again, the correlation wasn’t as strong as the correlation between this year’s attendance and last year’s, and in one year, 2024, it was tiny.

Correlations over the years for individual teams

I also wondered how the correlation between this year’s attendance and this year’s payroll, or this year’s on-field success, would look if you looked at just a single team over a period of years. Like the Phillies.

So I put in Phillies data on attendance and winning percentage since 1901, and payroll since 1998 (which is how far back the website I found went). To my surprise, the attendance (as compared to the league average) correlates  very strongly with winning percentage, far more strongly looking at one team for 125 years than looking at 30 teams over one year. I shortened it to just the time since 1998, and the correlation is just as strong. The correlation with payroll over the last 27 years is almost as strong, and is comparable to the correlation for the 30 teams for a single year.

Does this mean the Phillies’ fans are just fair-weather fans? No more than any others.

I thought a good comparison would be to look at a team that, like the Phillies has had some very good teams  and some very bad teams, but unlike the Phillies, ranks near the bottom of the majors in attendance each year. I took the Tampa Bay Rays, who played the Phillies in the 2008 World Series, and, like the Phillies, have lost a World Series this decade and, like, the Phillies, have also had some poor to miserable teams. For the Rays, the correlation between attendance and winning percentage is even higher than for the Phillies, but the correlation between attendance and payroll is very slight. There are two big differences between the Rays and the Phillies. One is that the Rays have learned to field excellent teams with very low payrolls, so jumps in payroll are, as often as not, a mistake. The other is that the Rays play in Florida, where people just don’t go to baseball games very often, so it’s hard to imagine them generating the attendance to generate the budget to field a high-paid team.