Vigil for Life? I'm going with 7,000

If imitation is the sincerest form of flattery, I should probably be feeling very proud of myself at the moment.

You remember how a couple of months back I strained my eyes to exhaustion by counting the marchers at the ‘March For Choice’, reaching a total of 850 or so marchers, which tied in pretty neatly with my initial estimate, a second estimate based on how much space Roman armies used to take up – because, after all, counting crowds is part of my training – and what people on the march had said about how much space the march took up, and the figures first reported by the Guards and the Irish Times, before they were pressurised into changing their numbers?

Well, somebody’s tried a similar kind of thing with Tuesday’s ‘Vigil for Life’, one Andrew Flood, writing on the Workers Solidarity Movement website, having valiantly taken up the gauntlet to count Tuesday's numbers, and, after dotting away on a 19-second video, having counted 1029 people at the vigil.

Andrew says he’s using the same method as me, but he’s not, and that should be obvious; I used three methods, none of which was without problems but all of which gave results in the same range, and for one of them I relied on information given by pro-choice friends who were on the march.

There is a superficial similarity between Andrew’s method and one of mine, as both of us took screenshots from a video and dotted heads as we counted them, but they’re substantively different. It’s ludicrous for him – or anyone else – to say we used the same system.

I counted based on video footage taken close up in daylight from a single inspection point near the end of the march, this being pretty close to best practice when counting moving crowds, whereas Andrew counted based on low-resolution video footage taken from a distance, at night, with such foreshortening that the rear two thirds of the vigil are basically indistinguishable.

This, you can be sure, is generally not how stationary crowds are counted. Still, I think Andrew deserves some credit for trying. It’s tedious work, and isn’t easy on the auld eyes.

1029? Bumped up generously to 2000? Sounds possible...

“Now obviously,” he says, “this is a broad estimate... it’s certainly not the 10,000 claimed by Youth Defence.  Not the 8,000 claimed by the Pro-Life Campaign and faithfully reported by RTE after some pressure.  It’s not even the 4,000 first reported by RTE.  The initial Garda estimate of 2,000 sounds more reasonable, allowing for a bit of coming and going and a few people being out of shot. Sprinkle on some wishful thinking and the fact that people had been visibly spread out to take up lots of space and you might get a reasonable Garda going for 2,000.”

On first reading this I was quite impressed. After knackering myself counting a crowd close-up in daylight, there was no way I’d have tried to do the same thing with a swooping long-range camera in the dark. Indeed, I’d looked at that same 19-second video before Andrew had done so, and assumed it wouldn’t be possible to count that crowd. It wouldn’t, I thought, even have been possible to estimate.

I’d seen somebody claiming there were only 300 there; I’d also heard that Youth Defence were claiming 10,000, which struck me as daft. 2,000 sounded quite plausible the way these things go. And, in truth, that’d not be a bad number, for the afternoon of a cold weekday, and everyone still at work. Still, I thought, I really wish people wouldn’t bump up their own numbers. 10,000 seemed almost as absurd as Ivana Bacik’s fantasy 5,000 at her March for Choice. Should’t we at least try to be honest?

And yet...

And then, after a few moments, I watched the 19-second video again, with the screen enlarged, freeze-framing as I went. I didn’t keep the stills, as with the one I’d worked from at the start of October. I just paused, and looked, and showed them to a (bored) friend, and very quickly concluded that this couldn’t be done.

Yes, I could see Andrew has taken four screenshots – as compared to the twenty I used for the ‘March for Choice’ – and dotted away, and done so ingeniously, applying red dots to people holding candles and green ones to people without them, but looking at this critically, I just didn’t think this was possible.

It was dark. The camera moves. It’s shot from a distance. There’s serious foreshortening with the crowd. The video’s low resolution.

Buses...

Still, it seemed an honest attempt at this, and I thought credit was due to him for trying. I only started to get uneasy when I heard somebody saying that about 2,000 people had come up to the vigil from around the country on buses they’d hired. Was that credible?

Well, a friend said that Youth Defence and the Pro Life Campaign people had told him that their local chapters had organised buses for 1,400 and 1,200 people respectively. That’d give 2,600 who’d come up in buses from the country, leaving aside people who’d made their own way there and those from Dublin and roundabout. Allowing for some double-counting, 2,000 sounded plausible.

What’s more, other data seemed to support this, as I heard of one person having personally arranged four buses from Galway, and of two buses having come from Monaghan, which has a lower population than all bar four Irish counties. More populous counties, you might think, would send more. Still, modern coaches take 55 people, while minibuses take 15-20, so even if we low-balled our figures, and assumed a coach and a small minibus from each county outside Dublin, that’d give us 2,170 people.

2,170 people excluding those who’d come under their own steam, and people from Dublin itself. Is a figure lower than 3,000 even remotely credible?

Common sense seemed to give serious grounds for doubting Andrew’s numbers anyway, so I decided to think about these figures again.

Crowd-counting 101...

The basic rule for crowd estimation is a straightforward equation based on space and density; it gets more complicated with moving crowds like ‘March for Choice’, but when a crowd’s standing still, as at a peaceful vigil, it’s pretty simple.

“There are two ways of estimating numbers,” says Nigel Stanley, head of TUC campaigns, on the BBC website, “One is drawn from having a rough idea in advance of the capacity of streets in the area where people gathered for the start of the march.”

To work this out you need precise maps with street measurements, multiplying width by length to get the area, which you’ll remember from school. Then you apply a rule of thumb that three people per square metre is comfortable and four is like a rock concert – the rule can vary between cultures, because Americans, for instance, tend to need lots of personal space, but this works well for British and Irish people. It’s important to keep in mind that density tends not to be uniform, and that crowds tend to be more compressed in some places, less compressed in others.

When dealing with marches you have to get stuck in and actually count samples and work out how long it takes people to pass certain inspection points – two or more for preference, but you do what you have to –  but with a stationary crowd you can and generally should just work with area and density.

So the professional crowd counters say, anyway, and a variant of that is my normal practice when dealing with battle accounts; it’s hard to read a field unless you’ve a good idea how many men fit on it.

Figuring out the perimeter...

So, how much space did the Vigil for Life take up? Well, if you’d selectively read the Twitter feed at the time, you could be forgiven for thinking there was hardly anybody there at all. 

“Looks like crowd at #vigil4life tiny fraction of crowd at any of the #savita protests despite national well funded mobilisation by bigots,” said @WSMIreland, adding “tricky to estimated size of #vigil4life as view obscured by expensive giant video screen & fancy PA but maybe a little over 300.”

“Tiny numbers at #vigil4life somewhat magnified by photos not revealing entire centre street is speakers platform,” it continued, proclaiming, “Pictures from scene suggest #vigil4life in hundreds @drivetimerte claims of 1000’s nuts,” and retweeting “They’re pipping cheers through speakers!! I'm at #vigil4life this is nuts. About 4-500 tops”.

Workers Solidarity Movement, then, seemed pretty sure at the time that there were only a few hundred there. Perhaps as few as 300. Certainly not the 1029 Andrew himself would subsequently count, let alone the 2,000 he was eventually to hazard, or the 5,000 Nell McCafferty – no pro-lifer she! – estimated on Facebook after having a gander on the spot.

@WSMIreland posted a few interesting pictures along the way. This one, taken from the Stephen’s Green end of Kildare Street is particularly good, as it shows the front end of the vigil, spilling forward from Molesworth Street across Kildare Street to the front gates of Leinster House.  

The WSMIreland caption to it is a bit baffling, though, decreeing “Pic showing just how small #vigil4life is compared to any for the #savita vigils held on Kildare st outside Dail.”

Well, it would show that if the Vigil had been along Kildare Street, but it wasn’t: it faced the Oireachtas buildings, rather than, say, the National Library and the National Museum, though it seems to have rather sprawled a bit on Kildare Street, if this rather colourful image is accurate. It’s almost as though WSM Ireland wanted to mislead people. Or just didn’t have a handle on what was going on. What’s your money on? Spectacular incompetence or egregious dishonesty? Either way, it doesn’t do a lot for their credibility on this issue.

I like this one too, which shows that the crowd stretched back to the junction of Molesworth Street and South Frederick Street. That’s useful. 

The Google Streetview version helps put things in perspective

A shame about the tweet linking to it saying “Picture showing spread out #vigil4life rally not even half filling Molesworth St. despite massive spending by bigots,” but there you have it.

I’m not quite sure why a determination to protect human life, and an unwillingness to have done unto other human beings what we would not like to have been done unto us – basic Golden Rule stuff, when you get down to it –  constitutes bigotry, but let’s leave that for now.

Well, now we have our crowd dimensions, if it not its density. The crowd on Molesworth Street stretched from wall to wall, taking up the footpaths rather than just the road, and as I’ve said, it went from the junction of South Frederick Street to the gates of Leinster House. Let’s pull it up on Google Maps, shall we?

And if we have the space, let's try for density...

I’m going to disregard the overspill sections on Kildare Street, because I have little idea how far the crowd spilled to left and right, and less idea how dense the crowd would have been where it wasn’t penned in my the edges of Molesworth Street. It’s best, I think, to treat it as a straightforward oblong.

Well, looking at the scale, and measuring it out like we used to do in geography class, it seems that the vigil must have taken up a space about 20 metres wide and 130 metres long, which – leaving aside the odd obstacle such as vans and screens – gives us 2,600 square metres. Now, according to the standard rule of thumb which sees three people comfortably together per square metre, this would give us 7,800 people at the vigil, or it would do if there were no obstacles.

If everybody was jammed together like at a gig, well, then we’d get 10,400, which supports Youth Defence’s claims, but I think we all know it wasn’t that crowded, so let’s stay with 7,800 tops, assuming a consistently comfortable crowd density.

But obviously we need to lower that because the area wasn’t entirely occupied by the crowd; looking at the aforementioned 19-second video, and a rather longer one made by Youth Defence, it looks as though they shared that 130 x 20 metre zone with a big screen, a speaking platform, and two or three vans.

None of these obstacles was all that big: if you look five seconds into the 19-second video, you’ll see that the platform can’t have been much more than 6 metres on any side, despite @WSMIreland’s claim that it took up the “entire centre street”. To be honest, looking at it in the Youth Defence video, I doubt any of its sides were more than 4 metres along, but let’s go mad and say that was 6 metres on each side, and go with, I dunno, 6 metres by four metres for the screen thing towards the back, and 5 metres by 2 metres for each of three vans. All told, that means 80 square metres need to be knocked off our total. 

2,520 square metres with three people per square metre at a consistently comfortable density, gives us 7,560. If the crowd had been packed like in a concert, that rises to 10,080 people, which ties in with Youth Defence’s claims, but judging by all the photos I’ve seen, and video footage too, I’m no more convinced by that now than I was on Tuesday; 10,000 has nice rhetorical value, but I think that figure’s wishful thinking more than anything.  7,560 sounds a lot more plausible to me, though it assumes an evenly packed crowd. 

Generally you should allow a 20pc margin of error on these things, according to the aforementioned Nigel Stanley, which suggests that there could have been anywhere between 6,000 and 9,000 people at the vigil.  Watson and Yip, however, reckon that if you’re confident on crowd density and area, you can apply a 10pc margin of error, which means somewhere between 6,804 and 8,316.

RTE’s eventual “several thousand” starts to sound reasonable when put that way, and the figure of 8,000, claimed by the Pro Life Campaign since Tuesday is certainly very possible. It’s not far off the mark, at any rate, and may in fact have been a conservative estimate.

6,000 sounds like a plausible minimum figure, then, but on balance, I’d go with 7,000, allowing that it might well have been rather more than that.

Foreshortening and density...

So how on earth can Andrew have gone so wrong, assuming he did so, which even the bus figures suggest he must have done?

An obvious starting point is that Andrew expected the numbers there to be low to begin with, as can be seen at how at 16:48 and 17:00 he’d retweeted the WSMIreland saying “Looks like crowd at #vigil4life tiny fraction of crowd at any of the #savita protests despite national well funded mobilisation by bigots” and  “tricky to estimated size of #vigil4life as view obscured by expensive giant video screen & fancy PA but maybe a little over 300 #savita”, before opining that the vigil’s organisers would have got a better turnout if they’d paid random people €50 each to attend.

Less than an hour or so later, in a thread which began with somebody else claiming that only about 300 were there, he said that pictures clearly showed that the crowd didn’t go halfway down Molesworth Street and were very spread out, with there being hundreds there, not thousands. At 19:33 he repeated the claim that Molesworth Street had been less than half full. Of course, even leaving aside the issue of density, it’s irrefutable that the crowd went as far as South Frederick Street, which is more than halfway down Molesworth Street.

Still, regardless of Andrew’s predisposition to number the vigil crowd in the hundreds, the video he worked from is of very poor quality, as I’ve said. It’s really hard to count from that, and when you contrast the video with the Google overhead shot, you start to see a massive problem. Look at the van which is on the border of his third and fourth panel. That van, which is three quarters of the way back in Andrew’s montage, was only a third of the way down the crowd-occupied portion of the street. I’ve marked its location on the overhead shot, locating its position with reference to buildings identifiable beside it. 

The video suffers from a terrible level of foreshortening, such that it’s about as problematic as a source can be, and appears to have led Andrew to say “really there is no way at all that 6,000 peopel were tucked away behind the white van.” 

Tucked away? Two thirds of the vigil’s length was behind that van! But then, Andrew doesn’t seem to have had any concept of that, claiming as he had done twice that evening that the vigil took up less than half of Molesworth Street, rather than about 60pc of it as well as a sizeable area on Kildare Street. 

Andrew’s count has 558 people in the three frames ahead of the van, and 471 in the frame behind it. Or, putting it another way, he has 558 people in the front third of the vigil, and 471 in the central and rear two thirds combined. Now that doesn’t sound very likely, does it?

Of course, it is just about conceivable that this could have happened, and if you work from a single source, as Andrew appears to have done, it can happen that you can get desperately misled into believing such highly improbable things. Single-source journalism, as the Irish Times seem to be discovering nowadays, is a mug’s game. The thing is, though, we don’t need to work from a single source. Here, for instance, is a still showing the back of the crowd, taken from that slightly longer Youth Defence video I mentioned earlier.

It’s a far better resolution than the video Andrew used, and gets in a lot closer, and shows us that the crowd was pretty dense at the back where Molesworth Street reaches South Frederick Street. Not crushed together like at the front rows at Slane, mind, but tightly packed for all that. Not surprising, really, given how cold it must have been.

Ah, but was it all like that? Perhaps it was dense at the very back, and maybe at the front, but nowhere else? Well, you know that van that Andrew used as a marker, the one that’s a third of the way back? Well, here’s a shot of the area around that.

We’re dealing with a pretty tight crowd, all told, pushed up right to the sides of the van. In fact, there seem to be about thirty people next to the van occupying as much space as the van. Again, it’s hard to count, given the darkness, but if we enlarge and squint and drop van-sized counting frames into position, it seems to back up the overall impression that this crowd is pretty compact, with roughly three people to every square metre. So far I’m not seeing anything to suggest that this was a sparse crowd.

It has to be said that Andrew, in his count, shows a screenshot where there’s what looks like a patch of spare ground. He presents this as evidence that the crowd was scattered. You can see it, about five seconds in. Watch the video. I could be very wrong, but I’m pretty sure this is the enormous “speakers’ platform” we keep hearing so much about. Yes, I know, not that enormous then; indeed, if you look at it on Andrew’s stitched together four-frame shot, you’ll see that it’s not that much wider than the speaker on it is tall!

In any case, it seems to be the only decent-sized gap in the crowd, and certainly shouldn’t be misrepresented as evidence that people at the vigil were standing metres apart.

Better to light a candle than curse the darkness...

Anyway, these last couple of screenshots from the video with higher resolution don’t merely give us an indication that the crowd density was high throughout. They also reveal that there were far more people at the vigil without candles than there were with candles. Seriously, try counting them. Take other screenshots from the Youth Defence video if you like, and if you can find good enough shots. You’ll find the same pattern across the vigil.

This is especially interesting, because Andrew’s count suggests that there were slightly more than twice as many people with candles as without. His figures give us 323 people without candles, as opposed to 706 with them. But it’s very clear from the better quality video footage that this ratio could hardly be more wrong.

Indeed, on the evening of the vigil, it seemed that @WSMIreland really didn’t grasp just how many people lacked candles, as can be seen in the caption to this picture saying “pic showing how #vigil4life organises have spread out crowd to make numbers appear greater at ground level”.

Leaving aside how poor quality this shot is, those dark patches clearly aren’t empty spaces. They’re people without candles, or people whose candles are blocked from the camera by their bodies.

That Andrew’s figures are hopelessly wrong can be seen by looking at that shot of the very end of the vigil. Andrew says he counted 167 people behind the screen, though having enlarged his pictures I can’t even see half that many dots. Do you reckon you can do better? On the clearer picture, I mean, not in Andrew’s foreshortened thing. Me, I gave up after 300, thinking that life is too short for this sort of thing.

That said, given that life is too short anyway, I’m glad more than 6,000 people took to Dublin’s dark streets on a cold weekday evening to say that we shouldn’t be cutting it short for others.

I'm going to let Andrew wrap it up, given he inspired this post. Very sensibly, he includes an all-important caveat in his post:

“A final word.  I actually don’t think the numbers mobilised have any importance as to whether women should have access to abortion in this country.  My position is simple, if one women wants an abortion she should be able to access it regardless of how large or small a majority of people agree with her choice. That is what the pro-choice position means - the choice is for each women and not anyone else to make.  However the political reality is that the arguments about numbers are going to be used to scare cowardly politicians so I’m not inclined to allow the crazy no abortion whatsoever forces that represent no more than 15% of the population these days away with anything.”

I half-agree with him. I don’t agree with his conclusions, or with his belief that only 15pc of the population are opposed to abortion, but I do agree with this: this isn’t a matter of numbers. Either we believe human life matters, and that without a right to life we can’t have any other rights, and that we shouldn’t deny others the opportunities we’ve had ourselves, or we don’t. Ultimately it comes down to that.

Like the girls said.