Saturday, August 14, 2010

Roissy's dating value test and Hot or Not score inflation

Last weekend I found Roissy's test designed to help men self-assess their dating value to women. I can't dispute that the guy knows his stuff, but by essentially giving each item an equal weighting, the scoring scale is less useful than it could potentially be. So on the basis of physical attractiveness, you're a 1, but the Hollister model standing next to you, who is a 10, tells the woman he's talking to that he has to use the restroom, while you just turn away from the one you're shooting it with in mid-conversation and go. According to the Roissy scale, all other things equal, the two of you are equally desirable in female eyes.

That is absurd, of course. Digressing, I can't believe simply walking away to use the restroom in mid-conversation is ever your best move--would it not be taken as a response to anxiety, the need for a private moment to regain your composure? I've actually been in that exact situation fairly recently. I told her to "wait here", and then went to relieve myself. When I came back, her mild curiosity and expectancy was met with an inconspicuous but relieved "whew" as I was a few paces away and then "okay, so you were telling me about your visit to Niagra Falls..." The response was good.

I've been a prickly skeptic of the game phenomenon not because I think it is fundamentally flawed, but because it is oversold. Confidence, assertiveness, wittiness, smooth operating, creating sexual tension both verbally and non-verbally, and everything else involved in game, ceteris paribus, raise a man's desirability. Roissy gives an estimate of 1-3 points worth. That sounds reasonable if a bit ambitious on the high-end, the gain depending on one's level of confidence to begin with.

To the extent that I disagree with him, it's in not seeing how he squares this with his ranking of game as being of greater importance than physical attractiveness. In addition to the restroom question, there are eight other scenario situations. Thus it appears that game is nearly an order of magnitude as important as physical attractiveness is for men. If you're doppleganger of Robert Pattinson, unless you are an extraordinarily agoraphobic mess, you're not going to have trouble running circles around witty, self-confident guys with crooked, crisco-infused faces and gorilla-hairy arms.

Tangentially, to gauge one's level of attractiveness, he suggests using the site I created a profile there a few days ago and began rating people. I was immediately struck by how inflated the scores are. After giving someone a rating, you are shown that person's average score from all ratings she's received up to that point. After rendering judgment on 100 or so ladies, I was without exception always harsher than the stated average, but the mean score I dispensed was probably 6.0-6.5, which is presumably a bit overly generous (it's a scale of 1-10, after all).

It seems as though the site's scale as it actually exists is 5-10, not 1-10. I googled "hot or not score inflation" and the like to try to figure out if there are bots at work or the site intentionally filters out low scores but didn't turn anything insightful up, only a few people making similar observations. The picture I posted has now accumulated 52 ratings for an average of 9.9 and the caption "You are hotter than 99% of the men on this site!" It's a decent though candid shot taken by a friend, not dressed up by a professional photographer or anything, and I wasn't hit by the ugly stick as many times as I could've been in the womb, but I'm not model-caliber. Does anyone know why HotOrNot scores are so inflated? The women who spend time rating male photos surely tend to be on the lower end of the looks scale, but that alone is unlikely to be a full explanation.

Since this is already dripping in self-indulgence, AE's responses and score on the Roissy test:

1. Age: 26-34 yo (+1)
2. Height: 6' to 6'4" (+1)
3. BMI: 24.1 to 27.0 (0)
4. Bench: 81% to 170% of body weight (+1)
5. Hairline: Full, under 35 yo (0)
6. Income: Over $70k, under 40 yo (+1)
7. Car: Have one, over 21 yo (0)
8. Looks: HotOrNot aside, I'm going to say 5-7, not 8-10 (0)
9. Sports: Played leading role (+1)
10. Occupation: Neutral status (0)
11. Friends: 20+ (+1)
12. Friends via internet: over 2 (a proud -1, thanks to those who comment regularly here!)
13. Last house party: Less than a year ago (0)
14. Funny: Nearly everyone who knows me (+1)
15: IQ: 130 to 145 (0)
16. At party: I approach first (+1)
17. Fight: Never been in one (0)
18: Arrested: No (0)

In the interest of space, just the scores for the scenario questions:

19. +1
20. 0
21. +1
22. 0
23. -1
24. +1
25. +1
26. 0
27. +1

+11, which means I don't even get a label. I've not spent much time in bars, so I'm not sure how accurate the description is.


Jokah Macpherson said...

"+11, which means I don't even get a label."

Greata beta perhaps?

kmr said...

I was curious whether I'd think Hot or Not's scores are inflated.

This was the third person they gave me to rate. I gave her a 3. Her average: 9.6


B Lode said...

Me, I was disappointed that I didn't get the highest rating. I'd really like stuff to stick to me the way guano sticks to car roofs.

Roissy's style is too off-putting and the test isn't designed for extreme introverts, who he'd qualify as extreme betas even if they have a lot more kids than him.

So, anyway, I consider him a distraction on the whole. His "diversity + proximity" equation is pretty good though.

Nick said...

I initially theorized that the female scores were inflated due to the influence of other women, i.e. women like to imagine the average woman is better looking than she is. And I imagine straight women are far more likely to rate women than straight men are to rate men, so this could create a disparity.

But then I looked at maybe 20 males pictures, and the scores were equally if not more inflated, as your particular experience suggested. The average had to be in the high 8s (Could there be a gay influence? I dunno what it would be?) There was even a classic lesser beta, meek and average and slumpy shouldered, and he still got an 8.6. I don't purport to be much of a judge of male looks, but this was still ridiculous. (Or maybe we're all adonises! I feel so much better about myself! Oh wait, if we're all adonises it doesn't much matter if I'm one too. Dammit.)

OneSTDV said...

Most people don't actually rate seriously on HotOrNot. Most people just go there to peruse the pics and click 10 for everyone just to get through them more quickly.

Audacious Epigone said...


Exactly. Even if you cut some slack based on apparent/stated age, as I do, there is no way she was ever in the 9-10 range.

B Lode,

Do you have a link discussing the equation you're referencing? I don't keep up with everything Roissy puts out to the extent I'd ideally like to.


Well, I guess sites like Hot or Not should be good for game in general, then, because it will tend to make guys feel more confident, which will mean more success with women, who will simply give it up more than they do now. Right?


That certainly sounds plausible to me, although I wonder why the default rating is always a 10. I'd think some generally misantrhopic people would instead always choose 1.

OneSTDV said...

Actually come to think of it, that seems more plausible. The Internet is a mean place after all.

Underachiever said...

I noticed the rating inflation too.

B Lode said...

Do you have a link discussing the equation you're referencing? I don't keep up with everything Roissy puts out to the extent I'd ideally like to. - Audacious Epigone

Oh, I think it is from here. Just aphorisms, more than a serious equation.

"proximity + diversity = war"

OneSTDV said...

I don't keep up with everything Roissy puts out to the extent I'd ideally like to.

Roissy is of course one of my daily reads. I don't read his site for the Game tips as I think once the basics are known that's sufficient.

I read the blog primarily because he provides insight that, while I fully accept it, still astounds me concerning the female sexual psyche.

rob said...

And here I was all ego-inflated over my 8.6

I wondered about rating inflation too, and couldn't find much quantitative. Hotornot used to post both their adjusted rating of the picture and the distribution of actual ratings. Last I logged on, no distribution. They also give percentile ratings where the 0?-10 score*10 is basically the percentile. My 8.6 is supposedly hotter than 84% of dudes on there. If I were that attractive, wouldn't I know?

However inflated the scores are, they seem precise. I posted the same pic twice to get independent ratings, after ~40 and ~200 ratings, one's an 8.6 and the other is 8.7. At least they keep their lies straight.

FuturePundit said...


I had the same reaction to HotOrNot a few months back. I was very disappointed. I do not understand why ugly chicks get such high ratings. Do people really have such low standards?

We need a site that does a relative rating system where people get showed 2 or more pictures and they click on them in the order of attractiveness. Even 2 pictures at a time with a click on the more attractive one would do the trick.

Maybe I should create such a site.

Audacious Epigone said...


Has that always been the case, though, or has the inflation become steadily worse? Are there bots at work giving out lots of 10s (for whatever purpose)? If your scores were so consistent, that would seem plausible.


Yes, that would be far more informative. It could be set up in the same way competitive ladder systems are--your pic goes head-to-head with someone else's, and the winner moves up while the loser moves down (with K contingent upon each person's rating going in). Have a winner declared after something like 200 ratings. A vast online looks hierarchy would soon emerge. Intriguing!

rob said...

Oh, I thought of another thing. They may be age-normed. I think the category I'm in is 30-40 or something like that. Maybe I'm just hotter than 40 year olds who let themselves go. Out of morbid curiousity, I've added a different picture. I wonder if the results will change.

Having such low self-esteem that I put my photo on hot or not should take a good 5 points any off any holistic measure of my attractiveness.

Steve Sailer said...

This reminds me of the Doak Scale in golf course architecture criticism. Tom Doak, then a golf course critic, now a golf course architect, invented a 0 to 10 scale in which the average golf course would be a 3. A 5 on the Doak scale would be in the top 10% or 20% in the world.

Steve Sailer said...

Here's Tom Doak's description:

The Doak Scale

0. A course so contrived and unnatural that it may poison your mind, one I cannot recommend under any circumstances. Reserved for courses that wasted ridiculous sums of money in their construction, and probably shouldn’t have been built in the first place.

1. A very basic golf course, with clear architectural malpractice and/or poor maintenance. Avoid even if you’re desperate for a game.

2. A mediocre golf course with little or no architectural interest, but nothing really horrible. As my friend Dave Richards summed up: “Play it in a scramble, and drink a lot of beer.”

3. About the level of the average golf course in the world. (Since I don’t go out of my way to see average courses, my scale is deliberately skewed to split hairs among the good, the better and the best.)

4. A modestly interesting course, with a couple of distinctive holes among the 18, or at least some scenic interest and decent golf. Also reserved for some very good courses that are much too short and narrow to provide sufficient challenge for accomplished golfers.

5. Well above the average golf course, but the middle of my scale. A good course to choose if you’re in the vicinity and looking for a game, but don’t spend another day away from home to see it, unless your home is in Alaska.

6. A very good course, definitely worth a game if you’re in town, but not necessarily worth a special trip to see. It shouldn’t disappoint you.

7. An excellent course, worth checking out if you get anywhere within 100 miles. You can expect to find soundly designed, interesting holes, good course conditioning and a pretty setting, if not necessarily anything unique to the world of golf.

8. One of the very best courses in its region (although there are more 8s in some places, and none in others), and worth a special trip to see. Could have some drawbacks, but these will clearly be spelled out, and it will make up for them with something really special in addition to the generally excellent layout.

9. An outstanding course—certainly one of the best in the world—with no weaknesses in regard to condition, length or poor holes. You should see this course sometime in your life.

10. Nearly perfect; if you skipped even one hole, you would miss something worth seeing. If you haven’t seen all the courses in this category, you don’t know how good golf architecture can get. Call your travel agent—immediately.

Steve Sailer said...

The point I wanted to make about the Doak Scale is that before Doak invented his scale to add rigor and high standards to golf course criticism, most golfers would have ranked the average golf course about a 6 or 7 instead of Doak's 3.

Why? Golfers like golf courses. They're nicer looking to them than parking lots or soybean fields.

It's the same with guys who rate girls' pictures. They like girls' pictures.

rob said...

I am now totally convinced that hot or not is broken beyond repair. My new pic is rated 9.8, and that is not true on this planet or on any other. Even if beauty were skin deep, my sould is far uglier than 9.8.

Anonymous said...

So much depends, anyway, on how hard you try to get a good pic.