In an interesting post Sean Lovelace suggests that the skewed stats at Duotrope make it a much less valuable resource than it should be. And this may sound rather slow of me, but erm, yeah. Of course. See, I was advised to use Duotrope by a few more experienced writers, and I joined up and browsed. I personally would prefer not to be published at "easy" places that appear to accept anything and everything, so I was looking for the ones that have lower acceptance rates, and then reading their zines, seeing if I thought my work would fit, keeping an eye on the market. But I rarely updated my subs file there,(actually, I think I did once) instead I write them all in a rather fetching notebook, tres old skool of me. Then I stopped looking, preferring instead to read other writers blogs and follow their recommendations, reading work I admire and aiming to be as good as that. Sean points out that saying a place has a 50 per cent acceptance rate only works if erm, like, a hundred per cent of people who sub report what happens to their subs.
Hmmm.
Thoughts?
EDIT - I have had a response from someone calling themselves thesingularitysucks in which they explain very clearly why Mr Lovelace's suggestion is incorrect, so it's well worth anyone interested reading through the comments.
4 days ago
12 comments:
I do think Sean is a bit harsh! Of course Duotrope stats are not 100% objective - but what are statistics? They are not some objective measure of anything....They are simply a guide, and I am sure that's what Duotrope will tell you. Anyone who believes them to be the gospel truth is, frankly, less than mathematically talented. Speaking as someone who really appreciates what Duotrope is trying to do, the real answer is for as many people to report to them as possible.
I track all my submissions in Duotrope's sub tracker, which is very useful - mainly for response times not for chances of acceptance. This isn't a game that can be played with statistics - whether an editor chooses to publish your story or not is totally subjective and no-one can say otherwise. But if I see from Duotrope's Recent Updates that in the past week 4 people have received rejections from X magazine, and I haven't yet received a rejection from X, I can take a little solace from the fact that they might like my story a little better and are hanging on to it for another day. I can also see when a publication begins responding to submissions, rather than be left wondering when and if I will get a response. That, to me, is far more valuable than "are they likely to accept my story because they have a 90% acceptance rate".
I think the problem I had is partly due to stupidity, I could not get my tracker to do what I wanted it to, so gave up. I think that duotrope could be a brilliant thing, but do wonder how many dedicated users there are. Still, if it suits you that's ace.
Try again - they have been upgrading the site so maybe it wasn't working properly before. The sub tracker is a little unwieldy, takes a while to do things, but I find it useful!
My tuppence worth: I use it and love it. I'm not sure about the accuracy but when I only started to submit (um, all of a few months ago) it helped me find markets I never would have. I use the tracker but then I also have a geeky spreadsheet with colour coding. At the end of the day it's a free service with, if nothing else, a list of possible literary markets and their websites and for that alone it's so helpful to people like me who aren't familiar with the market yet.
I second everything Kerry says, but would just like to point out that it's not really free, they do ask for donations, and I donate a small amount when I can, since they are providing such a great service. Even a dollar every now and then would help!
Hi Kerry. Oh yeah, it's very cool as a list of possible places to sub, although I think that some of the info is out of date. And I don't mean to knock a free service. I was just interested in what Sean had to say, I hadn't thought about it like that.
By the way, have the comments gone really small for everyone or just me?
Hi Tania. Ah, yes, they do code reds and so on don't they to show if they need money or not? Who is responsible for it actually? Does anyone know who runs it?
I think Duotrope is really good as a resource.
Sure, the stats are totally reliant on regular (and honest) updating on the part of the users, and we all know that aint going to happen! But yopu can still cross check with different magz... one that has poor beginnery work will have a higher acceptance rate than one with a higher standard.
and you gradually get to see where writers you recognise/admire get published...
and I have to say, without tit I wouldnt know about half of the places I like reading, too.
> Sean points out that saying a place
> has a 50 per cent acceptance rate
> only works if erm, like, a hundred per
> cent of people who sub report what
> happens to their subs.
If Sean says that then Sean is flat-out wrong. If he were right all statistics would become completely unreliable, the core of the scientific method would be overturned, and our civilization would collapse as supposedly established scientific principles suddenly proved to be false.
Electronics, for instance, isn't built upon observations of a 100% of electrons. It's built upon observations of quite a few electrons over quite a bit of time, but we can't watch all of them all the time. Nevertheless, the theory holds broadly true, and sufficiently true to be useful.
Statistics work because the majority of a group, be it writers or electrons, tends to behave the same as each other. Yes, there are always a few 'crazies' (there are certainly crazy writers and I wouldn't be surprised if there were crazy electrons too) but the more data-points you have, the clearer idea you have of how things behave.
Consider a market with a 50% success rate. there's going to be a few 'crazies' among the people submitting to the market who either get 100% success (good crazy) or 0% success (bad crazy). If we just took a reading from a single person, we'd be in danger of that person being a 'crazy' of either type, and thus giving us an incorrect statistic.
However, the more readings we take, the more accurate our figures become, as the 'crazies' are swamped out by the 'normals'. Yes, if we sampled 100% of people, we'd get the most accurate figure possible. However, if 10,000 people submit to a market, and we sample 500 of them, we'll probably get a fairly accurate figure, as most of the people in the sample will be 'normals', unless for some reason we've managed to take a sample that has a higher than normal level of 'crazies'. Indeed, if we get all normals in our sample (which is far more likely to happen than getting all 'crazies', because there are more 'normals' than 'crazies' in most situations) then we should get pretty much the same figure as we'd get if we sampled 100%.
There are of course other factors with duotrope. For instance, some people may be far more inclined to report their successes than their failures. This will skew the figures slightly towards showing markets as having a higher acceptance rate than they do. However, one would not expect this kind of behavior to effect some markets more than others, one would expect it to be fairly evenly spread across all markets, as presumably the people who do this submit to many markets. Hence, in terms of comparisons between markets, they are all slightly too high acceptance wise, and the comparisons that 'this market accepts more than that market' should still hold true.
Of course, there are many markets on duotrope that have very few data-points, and thus have very skewed figures. Few people have submitted to them, so the figures that duotrope have are likely to be 'crazy' figures. Duotrope knows this and generally puts a warning on the pages for these markets 'we have not had enough responses to give an accurate figure'. However, there are a lot of markets where a reasonably accurate figure if available.
To pick out markets where the figures are skewed due to low response rates and declare that all figures on duotrope are useless is to throw the baby out with the bathwater. Einstein did something like this when he showed that there were certain situations in which Newtonian physics did not apply, but this didn't make Newtonian physics useless, and nor did it cause the things built on Newtonian physics (like steam engines) to suddenly stop working! Hence, though someone using duotrope must be aware that the figures they see are not the gospel truth, nevertheless they will continue to be useful most of the time, in those cases of markets where there has been quite a number of responses reported.
(I had to break this post up into chunks because blogspot got all fascist on my ass about it being "Too long")
I use duotrope, I must admit I don't generally care too much about the acceptance ratios of a market, being far more interested in response time and pay. However, there seems to be a lot of anti-duotrope feeling about the web these days, and I wonder why? After all, what's the alternative? No one is ever going to be able to poll 100% of submissions to fiction markets, so the choice seems to be between having statistics with some issues, or having no statistics at all. I report all my submissions (giving me something like a 7% acceptance rate :-( ) I think quite a lot of others do too. Duotrope may not be perfect, but I find it the most useful tool that's on the web for writers. Sean says in his post that most of the markets on duotrope are closed, but Sean must be doing something wrong, because whenever I run a search on duotrope I am confronted with a list of currently OPEN markets to which I submit.
Without duotrope I guess there would be Ralan, but between these two I've found duotrope to be by far the more useful.
Thank you so much for your post thesingularitysucks. Lots to think about and you have me convinced. I am rubbish with numbers but you've explained it very clearly. I'm going to update my post and ask anyone who reads it to make sure they check out your comment. Cheers.
Post a Comment