Poll: Half of College Football Players Say They Have a Gay Teammate


In ESPN magazine's college football preview, Bruce Feldman and Ryan Hockensmith survey 85 current college players about the state of their game. One of the questions players were asked was if they had any gay teammates:

"Almost half of those surveyed (49.4%, to be exact) said yes, they believe they have at least one gay teammate. In the Pac-10, 70% of those surveyed said yes."


  1. Focus says

    This survey is meaningless. Only 85 players were surveyed out of the tens of thousands in division I alone. The sample size is tinier than the steroid-shrunk testicles some of those players are sporting.

  2. B says

    There are hundreds of players on each team and there are hunreds of teams in Div. 1 alone…a survey of 85 doesn’t even total half of one team. This seems to be a pretty meaningless survey.

  3. says

    n its infancy, a news gathering was primitive by today’s standards. Printed news had to be phoned in to a newsroom or brought there by a reporter, where it was typed and either transmitted over wire services or edited and manually set in type along with other news stories for a specific edition. Today, the term “Breaking News” has become trite as broadcast and cable news services use live satellite technology to bring current events into consumers’ homes live as they happen. Events that used to take hours or days to become common knowledge in towns or in nations are fed instantaneously to consumers via radio, television, cell phones, and the Internet http://1-news.net/

  4. Seth says

    Well, 85 is way too small of a sample to tell anything, but I think 50% is probably an accurate-ish number. I work with a Div. 1 college team and I’d say that, from my experience at least, the locker room is getting a lot more gay friendly.

  5. Sean R says

    So… they knew 1 in 10 are gay because the other 9/10 won’t “oblige” when asked?

    It is an interesting article but a missed opportunity to get beyond rumour/ counter-rumour. It is clear many athletes stay closeted because of homophobia and threats to a some-times lucrative career. But if this survey was genuinely anonymous, why wasn’t it asked simply “are you gay/ bi/ hetero?” and try to get a more precise reading? (ok yup people lie in surveys, etc. but its a bit more scientific than ‘do you know any gays in the village, but it would do more/ challenge the silences around sexuality to ask people to identify their sexual orientation rather than cop out with a ‘do you know…’)

  6. Mike says

    Actually, even if there were 30000 players, a sample of ~85 would still give you a confidence interval of 11 and a confidence level of 95%; so you could be 95% certain that between 38.4% and 60.4% of the players would answer the same way.

  7. Pee Town says

    I have volunteered repeatedly as water boy. One of my many qualifications is my friendly agnosticism to a player’s sexuality. If I’m in the locker room, I’m an equal opportunity fluffer.

  8. Chitown Kev says

    Thankfully, they didn’t put Tim Tebow on the cover. Of course, if you follow college football (and I do) you do hear rumors. Tim Tebow (good Christian boy that he is) for one, gets a lot of anti-gay slurs thrown at him from the opposing teams fans.

    Colt McCoy (who is adorable) not so much

  9. GregV says

    @Sean R:
    You’re right that there are a lot of variations on the gay question that would have been more telling and more interesting. Asking whether they “think” there is probably a gay team member would yield a different result than asking whether they are aware of a gay member or whether they are gay themselves.
    There really is no survey that can provide enough assurance of “anonymity” to bring out of the closet anyone who is deeply in it, though.
    I remember being in those situations when I was younger and didn’t feel safe being out. If two players are dating and want to keep it on the “down low” around fellow players, a published result saying that there is “a total of two members who are gay” on such-and-such team would end rumors and pretty much out them to anyone who was wondering.
    I remember when a teacher said our responses to a sexuality survey would be completely anonymous, and I was thinking, “What if the teacher sees what color my pen is? Or what if my answers to questions about race and subjects studied were to give me away… I might never be safe in the hallways again,” so I checked “heterosexual.”
    So any “anonymous” survey on sexuality is always an underestimate, and always only indicates the numbers of respondents who are at a certain degree of being “out.”

  10. says

    Thank you, Mike! I’m sick of hearing total innumerates declaim on what constitutes a proper sample size. 85 is plenty! Usually these sample-size queens come from the Left, but when the trollish Right gets onto Towleroad, the same misology rears its ugly head.

  11. Mike says

    A few points as a follow up to my previous comment:

    1) Thanks for the kind comments, and if you ever want to check statistics like this and don’t know how, you can use several online calculators, like this one:


    It’s easy!

    2) There are 120 NCAA Division I football teams, and the average college football team has 125 players, so that’s about 15,000 players. So, the actual confidence interval would be 10.6 at a 95% confidence level, so we can be 95% confident that between 38.8% and 60% would answer the same way.

    3) It’s a fucking poll by ESPN, not the final, published results of a clinical trial in a peer-reviewed medical journal, and shouldn’t be held to the same standards.

    4) I recently worked for seven years for a major university as the data manager for dozens of clinical trials and longitudinal studies, and if we ran across preliminary data that showed we could be 95% certain that a vitamin would cure a disease in 38.8%-60% of the general population, we would be *very* interested in doing more research— right after we double-checked our numbers and then wet our pants.

    As usual with numbers, it’s all in how you frame them.

  12. nic says

    in 2008 the sampling of likely voters was somewhere between 800 (the marist poll) and 3000 (reauters/cpan/zogby). they came to the same conclusion. irrespective of the tens of millions of voters who came to the polls.

    if idiots say that a sampling of 85 college football players is meaningless, perhaps they should question the scientific quantification of statics, rather than the size of the sampling.

    aside from that, i am happy that young, straight football players don’t seem to care much about playing with gay teammates.

  13. Yuki says

    I definitely like the idea that the football players seem to be fine with it.

    I see a lot of people getting mad at people complaining about the 85 people that were polled. I can let it slide since it’s not a scientific study or anything, but in even in high school science fair projects, you tend to need a minimum of 100 subjects; anything less than that isn’t as accurate, according to every teacher I’ve ever had.

  14. Brian says

    The survey means TowleRoaders probably want to know what schools are PAC TEN:

    Stanford University
    University of Southern California
    University of Washington
    University of California, Berkeley
    University of California, Los Angeles
    Washington State University
    The University of Arizona
    Arizona State University
    University of Oregon
    Oregon State University

    Have fun.

  15. nic says


    i guess your teachers need to do some research. as you said, the writers of this survey were not being scientific. nevertheless, challenge the formula; don’t aim at the sampling. what would your teachers have said about polling 500 likely voters and drawing conclusions about 165 million+ potential voters?

  16. Mike says


    In general the more data you have the more accurate your results, but there is nothing magical about having 100 data points. Prestigious, peer-reviewed medical research journals are filled with clinical trials and longitudinal studies with less than 100 subjects.

    A quick way to *estimate* the accuracy of a sample size is to take 1 divided by the square root of the sample size, so 1/the square root of 85 = 10.8%, which is close to the confidence level of 10.6% we found above.

    As you can see by using the above formula, the only thing that changes if we go from 100 subjects to 99 subjects is that the accuracy goes from 10% to 10.0503782%.

    If anyone really wanted to question this poll, they should focus on other issues besides the sample size, including randomness and bias. Even medial research professionals get these things wrong, because they are not statisticians, so they would consult with my department at the university and we would audit and modify their study designs as necessary. It is common for universities to have these departments for this purpose. To be clear, I am not a statistician, but worked with them for seven years while I was a software engineer and data manager in the statistical consultation and research department.

  17. Yuki says


    They most likely would have said it would need more, I should think. I believe I said “at least 100″, too… the more people you get for a survey, the more accurate it can be.

  18. nic says


    look it, if we were to randomly choose 85 males out of the entire u.s. male populace, and ask them if they know a gay man or woman, what do you think they would answer? “durr, i don’t know. i’m too busy picking my nose and fuckin’ my sister to care.” i don’t think so. likewise, your conclusion on this non-scientific survey.

    because you and others are averse to random sampling and cannot see the forest for the trees, then look at what MIKE and others have said about scientific polling.

    i am merely trying to explain the obvious. unless the 2 journalists specifically picked out homo-friendly college football players, then we must assume that nearly 50% of 85 collegiate football players are being untruthful. moreover, we must assume that 70% of the pac-tenners are liars. that makes no sense. meanwhile, i’m packing my bags for the pacific west to see what ‘poling’ i can do. yumm.

  19. Isabella Cane says

    This survey is meaningless, it asked if they think someone on their team is gay, there is no facts in the survey its just a matter of opinion, and even if it was true, the gay football memebers would most likely stay in the closet because they would be srutinized, and they would fear what their teammates would say or do, and the truth is until everyone is ready to look at the world open minded and without persecution of someones sexuality or even ethnicity, sex, etc. then there will still be those people who are gay and they will be in the closet, and people really need to get over the stereo-types, why dont they focus on all the steroids being used in football instead of going after the gays.

  20. Tigger says

    The PAC 10 is now the PAC 12 with the addition of:
    Univ of Colorado
    Univ of Utah

    That said, I want a hot Mormon football playing husband. Thx.