Last weekend, I was at a chess tournament and a fencing tournament. At the chess tournament, I could tell approximately how good someone is by the number after their name, but had no frame of reference for the fencing rating system.
At the stabbing competitions, some of the competitors have a rating, a letter after their name from E to A. The problem is, it doesn’t really mean anything to anyone who doesn’t already know about it. If someone has a B rating, does that mean they’re tough but beatable, easy, or you-should-run-the-other-way-and-scream?
I’m familiar with this problem from failing to explain what a particular number in a chess rating means, but I’m not used to being on this side of it. I had no idea how impressed to be at any point. So, in the “things that will probably interest precisely nobody else” category, I decided to try to come up with some way to convert between the USFA and USCF ratings, just so it makes sense to me. I have no idea if I succeeded.
Some cursory Googling turned up this forum post from the distant past, saying that about 2% of epee fencers have an A rating. (1.99% was the biggest number available. I should probably be more unsettled than I am by the contradictory numbers citing a different now-unavailable source, but I decided to use the bigger one and call it a margin of error.) The vast majority of fencers are unrated. In chess a rating is a prerequisite for all the good tournaments; in fencing you earn one.
The chess people have some conveniently available data, and from an official source, no less. (Oh yeah, these are just American organizations, if the “US”-es in the acronyms didn’t give that away.) This is out of date by about the same amount as the numbers I’m using for the fencing part.
Before comparing the numbers, I’ll get all the “I don’t know what I’m talking about” disclaimers out of the way first, and then go on to talk as if I’m definitely right. First: I don’t know what I’m talking about. I’m basically completely unfamiliar with the USFA rating system. Second: I’ll be pretty much assuming that whoever has the higher rating is better at their sport. This is only true on average. And I’ve heard that upsets are more common in fencing than in most sports, especially if you’re fencing epee. Third: If I told you that skill at fencing and skill at chess are as different as apples and oranges, you’d laugh in the face of my understatement. All I’m doing is comparing percentiles. Third and a Halfth: You can compare percentiles of anything. I’m using chess and fencing ratings, but it could just as easily be marathon time and number of blinks per minute. There’s no actually meaningful quantity being measured. Fourth: I don’t have access to the original source for the fencing numbers, so I hope nobody minds hearsay.
So, those are the reasons I don’t actually trust these calculations. With that out of the way, I am now completely right about everything and definitely have all the numbers I could possibly want.
I want to compare how good someone is at fencing to how good a chess player is, because the chess ratings are numbers that actually mean something to me. So, compare the percentiles. An A rating for an epee fencer means they’re in the top 2%. (1.99, but who’s counting.) For a chess player (use the non-scholastic column for more fair numbers), the top 2% would mean a rating a bit over 2200. So by my calculations, losing to an A-rated fencer is approximately as embarrassing as losing to a low-level chess master. (For most people, this means, “not very.”)
Now assume that all rated fencers are uniformly better than all unrated fencers. Then a B rating for a fencer corresponds to the top 5.15%, which is about the same percent as a USCF rating of 2000. That’s just about exactly expert level. Do the same thing with the other fencing ratings, and you get that C corresponds to 1900 level, what chess players call class A. (Just to be confusing.) And fencing rating D means a slightly lower-rated class A chess player, and E to a low class B. This means that there are several categories full of people who are better at fencing than I am at chess. My ego will not stand for this; fortunately it doesn’t have to.
In real life, of course, not all rated fencers are better than all unrated ones. At high levels it’s true, because anyone who can compete that well is probably experienced enough to have earned a rating. But at lower levels, it’s entirely plausible for someone to be good enough to have a rating but not have it yet. Let’s assume that instead of all unrated fencers being worse than all rated ones, they’re actually just as skilled with the same distribution. So if 10% of rated epee fencers have an A rating, we’ll take this to mean that 10% of all the competitors are that skill level.
Now an A-rated fencer is only in the 90th percentile, corresponding to not a chess master nor even an expert but somewhere in the middle of chess class A. For everything else, a fencer with a rating in some letter would be in a worse percentile than a chess player with the similar-sounding title. An E-rated fencer would be in the same percentile as a chess player rated 800. I’m not even sure that someone who played at that strength could beat non-chess-players reliably, and an E-rated fencer has done well in at least one fencing tournament full of fencers who fence.
The most accurate comparison will be somewhere in between the two, pretty heavily weighted toward the first estimate. (The one that’s good for fencers.) It’ll get even closer to the first estimate as the ratings get higher. A fencing rating of “A” probably does correspond more or less to chess master level, so I don’t feel too bad about losing that one bout five to zero in seventeen seconds.