Wednesday, April 30, 2008

Online Tax - Timewatch edition

Ancestry.com, continuing their programme of digitising government records (they have a deal in place with the UK's Public Records Office for Census data) have digitised US tax records from 1862 to 1918.

Want to know what Abraham Lincoln paid in 1864? $1,296 in tax from a salary of $25,000 - or something over $1/2 million in today's money.

200804301031.jpg

You can also see that Andrew Carnegie earned $12 million (or $84,000 in 1865). If you're looking for Mark Twain, remember his real name was Samuel Clemens.

The list of other data items they have compiled is enormous - this is just a short section from the list.

200804301035.jpg

Sourced from BusinesWeek.

Tuesday, April 29, 2008

Great Expectations

When you have a bright business idea, figuring out what the market size is - and what percentage of that you'll actually get to use your product - is pretty hard. In the dotcom days, the approach always seemed to be "well, there are 60 million people in the UK and 50% of them have dogs, and 50% of people are online, so we have 15 million people who can buy dog food online, so that's our market. And we think we'll get 20% of those in year 1, so 3 million customers in our first year". They call it top down analysis. And it leads to wildly inaccurate results.

At the beginning of the UK e-government programme, when the PM had said he wanted "100% of government online" there was some lengthy backchat about how it was the wrong target and what was needed was a take-up target. That would be how to manage it the nay-sayers said. Things went back and forth on the sidelines for quite some time - but plainly the PM had spoken. I make three comments here, reiterating what I've said several times on this blog:

1) the PM didn't need, in my humble view, to talk take-up here. He wanted to move the UK on and did so by declaring a vision that in a few presentations at the time I likened in simplicity to the "put a man on the moon, by the end of the decade, bring him back alive" speech. The PM said "get 100% online, do it in a citizen focused way, do it together, all joined up, do it by the end of 2005". Did it happen that way? Not really - but it wouldn't have happened any better, I think, if the PM had said, "get me 100% usage of the top 10 services."

2) the Treasury then took that big picture target and did actually start putting take-up targets on some departments. The Inland Revenue picked up a 50% target. Arbitrary? Certainly. Did they meet it? No. But that doesn't mean it wasn't a valid target - it was stretching, time-boxed, measurable and all of those good things. And, more importantly, people spent time trying to get it done.

3) One of the interesting effects of the 100% target was that when we took a look at plans for realising the target, they showed gradual delivery of services in 2001, 2002, 2003 and 2004 - and then a sudden ramping from perhaps 10% of service online to 100% of services online in the last year; in many cases, the ramp was in the last quarter of 2005. Lots of things went in the "too hard, think about it later" box and so, it was assumed, pixie dust would be sprinkled, magic would happen and services would go online. It took real work to develop a realistic implementation plan - but that plan was greatly helped by the absence of a debate about which services had to go online. When the answer is "all of them", it's only a question of order, not about whether.

On a recent trip to Paris I was excited by the Velibre scheme. Thousands of bikes spread throughout Paris that you can check in and check out when you need one. You can ride one from Avenue George V to Pont Alma, putting it in a rack at the end of your journey. When you want to go back to George V, you simply take another bike and return it to the original rack at the end. Or maybe you go back by Tube and don't have to worry about taking a bike back at all. A brilliant system.

What got me particularly was that the way you rent a bike operates at several levels of interface complexity:

  1. You have the full interface, with a big screen on a post, where you put your credit card in to set up an account from scratch and where all the rules are explained to you (in a variety of languages)
  2. A smaller screen on the reverse of the post where you can key in just your subscriber number and a bike number and off you go
  3. A card reader right by the bike where, if you have a long term subscription (so are local or in Paris often I imagine) you can just swipe the card and take the bike.

Back in the early days of direct.gov, we spent quite a bit of time thinking about whether there was a need for a tiered interface - one that dealt with expert users and beginner users quite differently. Various sites handle it now by giving you different entry points - so they have a cookie in your browser and if they detect it when you visit again, they skip all the instructions and go straight to the transaction. That's not quite the same as we envisioned - we wanted content to be greater or lesser based on how familiar you were with the process, how "expert" a user you were. I don't think we ever fathomed a way to do it in a way that made sense and when I look at most government sites around the world now, I don't see anyone else doing it, so perhaps it was a bad idea back then. So to see it implemented this way in Paris was a pleasant surprise. That said, I saw many people turn away from the Velibre console, baffled by it - as was I in the first place, because I looked at the second screen (on the reverse of the post) only and couldn't see where I was suppose to put my credit card.

I wondered how the business case for the Velibre system looked when they presented it, presumably to the Mayor of Paris. It would have been fun to listen in to.

"So we're going to buy 3,000 absolutely identical bikes. And we're going to build racks for them where cars normally park, so they'll be fewer parking spaces. And we're going to let people rent them. And if they use them for a 1/2 hour or less we're going to make it practically free. If they use them for longer, then we'll charge them a few euros, maybe 20 or 30 at most, for a whole day"

And then the questions would come

"But where will the cars we displace park? We don't let them park illegally now you see. And what about the bike stores - won't they complain about lost sales? And who's going to handle the service and maintenance? And what about vandalism? And what happens if someone gets a puncture on route? And won't that mean fewer people use public transport? But what about when it's raining? What would the liabilities be if someone were hurt"

I suspect all of those questions had really simple answers. More bikes on the street means less people using cars, means less need for spaces. Giving people the chance to ride a bike in the centre of town probably leads to increased sales for those who want to ride elsewhere, the bike shops probably saw business grow. There's a maintenance contract in place. Vandalism seems to have been a non-issue - I wonder whether there were cameras positioned near each rack, keeping an eye on them. And so on.

And market size? Who knows. I saw full racks and empty racks. I saw groups of 6 or 7 people coming together, all renting bikes and cycling off together. I saw tourists in 2s, 3s and 4s picking up bikes and taking off to see the city. I even saw a couple of guided tours where most of the people were using Velibre. Certainly if you sit at a table in a pavement cafe, perhaps by the Rue De Rivoli, within 10 minutes you'll see 30 or 40 bikes from Velibre pass by. From everything I saw, the metrics looked good - they'd taken a visionary punt with an idea that I can imagine took some tough selling; executed it brilliantly (sample of one, on one day) and had been rewarded with enormous take up and lots of very happy looking people cycling through Paris.

The proposition is simple and elegant though. Try a bike, take a look at our city, pay not very much money. Want to go to work by bike? Fine, take a bike, leave it near your office in a rack that we've handily positioned. Want to go out for a drink tonight? Fine, cycle to the bar, park the bike, leave it behind and get a tube or a taxi home. Absolute genius.

In e-government, we all had bigger than probably justifiable views on what take-up should be. I was probably the first dummy that thought a million self assessment registrations was viable within the first couple of years. In hindsight, the proposition wasn't simple and elegant enough. Visit our website, sign up to this, wait a couple of days, now fill in this complicated form using data pulled from all kinds of places that you've probably forgotten existed or never knew you had. No wonder take up was low to start with. It's risen dramatically since then - as has take-up for a variety of services.

There are still mountains to climb. I wonder whether any government, anywhere can claim 100% usage of its online service (i.e. 100% of the target market or, perhaps, to be a little fairer, 100% of the online part of the target market)?

But I look at Velibre and I'm pretty sure that 100% of its bicycles have been used and that, better still, it has created a market far bigger than its original expectations - and so had a variety of additional payoffs (dare I say healthier people, happier tourists, booming business for bike shops?).

So, be bold. Be very bold. A vision needs to capture the hearts and minds of people, inside and outside of your own organisation, government entity or department. If we had our time all again, I wouldn't change, for a second, the "get it all online" message.

Wednesday, April 23, 2008

Tag Cloud's a Crowd

This is the tag cloud from my blog, as parsed by tagcrowd.com. Oddly, the HTML that the site generates isn't quite right - it fails validation in ecto on almost every line. So this is a pasted image. For an e-government blog, I'm not doing as well as perhaps I should be. Must try harder.

200804231607.jpg

Poker Votes

voting.jpg

It's 2008 and it's time to vote again. For all sorts of things, be they pink, peach or yellow. Or Red, Blue or Gold for that matter. In 2001 when it was hotter than hot to think that everything would be on the Internet in no time at all, we all (well, most of us) imagined that voting would also have succumbed to the online wave by the end of 2005 (along with every other UK public sector service) and certainly, if not by then, by now. But today I'm looking at the postal voting form that arrived, handily, in the post.

I've got 4 different voting forms (there's a by election here too). I have the mayoral vote, the London Assembly one for my area, the LA one for London as a whole and a local councillor form. Pink, Yellow, Peach and White. I've also got 2 envelopes, both white but usefully named A and B. Perhaps this is why it can't be put on line - it's all too hard. No point paving the cow path, got to simplify the process and put that online.

I've also been looking at the VoteMatch site - if the mayoral candidates are all too similar for you, alleged corruption and clownery claims aside, it's worth taking a look at and 200804231231.jpg seeing if the vote you have in your mind matches the way the issues are addressd by the candidates. Assuming VoteMatch has the facts and the weightings right - and it could all be a clever stunt by Ken or Winston (but probably not Boris) then it really might surprise you.

It would be pretty cool if, at the end of the 25 questions, it said "do you want to vote for the candidate that most matched your preferences? Click here to enter your government gateway ID and vote." It is, in many ways, a great shame that there are so many blockers to achieving that (it would be a long list: linking the gateway ID to your actual vote, security issues, the risk of someone pressurising you to vote for their chosen candidate rather than yours and so on - there are ways out of all of those but the electronic voting process is pretty heavily tainted and I don't see anyone fixing the issues soon).

But I wondered whether there might be a better way to do it, harnessing both the excitement of a poker game (seemingly the fastest growing indoor activity after, well, after sitting watching TV) and a bit of Internet voting technology.

I'm a pretty terrible poker player but I like to play it, for small stakes at least (when you know you're not going to win there seems little point in playing Vegas stakes). The build up of the game is what fascinates me:

There's information in the middle that everyone has, and then gradually more information is made available that is both true and inevitable (more cards laid on the table) and potentially false and misleading (the chips that each player bets with).

The game unfolds a little at a time. Little is learned in the first hand, but after a few, you start to pretend that you know how the others are thinking. Your own bets get bigger as you get bolder (drunker?).

Money moves from player to player until, in the end, one or two players have most of the money and duel to see who comes out on top, winner takes all.

Now that sounds to me like an election.

If we compare the US primaries to our own Mayoral election, why wouldn't we do something similar for London?

Primaries in every borough. Spread over a few months, or even a few weeks. The candidates get the chance to persuade small sections of the total London population what they're going to do for them. The other boroughs get to watch with interest. They can let one candidate get an early lead and then rope them back in later. They can see what promises are made for, say, SW2 and compare them with promises made for SW6. Over the length of the campaign, more and more of the issues would be exposed and voted on; candidates would not win or lose in a big bang, they'd have a chance to tailor their policy to different areas and win votes on a more local basis. I imagine the actual borough councils would not be so impressed with these idea, after all it's they who run the boroughs, but what then is the point of a "London Mayor" if he can't interact at a local level as well as a macro-level.

So how would that work on the Internet

- Sign in, using your super-duper yet to be developed "I'm an online voter" (as opposed to a postal voter) token

- See where the votes cast so far stand. See who is up, who is down, who is fastest growing this week, this month, this hour, this minute

- Check the forums to see who is saying what about which candidates. Browse the press gallery. Check for the latest rumours

- Test your vote preference against votematch or any one of a dozen similar services that would be developed (perhaps as facebook widgets, although I probably hope not)

- Make your vote

- Check back next day, next week, next month and see how things are going. Is your candidate gaining, or losing. Do you want to pull your vote and tactically vote for someone else to prevent another candidate getting it? Of course, your vote alone isn't going to make the difference, but every vote counts. And maybe there's a tool there where you rally your friends to vote with you, kind of like establishing a proxy system

- In the countdown to the decision, the votes will move back and forth faster; maybe blocks of votes can be traded in return for the candidates agreeing a significant policy change. Maybe the losing candidates can donate their votes to another candidate who is ahead of them - tactical voting again. Maybe candidates can join together and form a coalition.

Pie in the sky I know, but so much potential. This would be a like an uber-rolling poll system like those conducted by yougov. Real people casting real votes day to day; maybe borough by borough, maybe in London as a whole.

On a related note, I was surprised to see this entry for Winston McKenzie when I googled his name. Winston is an ex-boxer from Croydon and so very loosely you could argue in the "demolition" business I suppose. The demolition-related words are in the source code of his site. I wonder whether whoever did his site reused a template from some other site?

winston.tiff

Tuesday, April 22, 2008

Forgive me Fartlek, for it has been a year since I last ran

Or how to recover from multiple meniscus cartilage tears - a guide that might help fellow sufferers

In 2006 I ran the London and New York Marathons. I managed respectable, sub 4 hour, times in each and met the goals that I'd set for myself and I raised a good total for Macmillan Cancer Relief. I set about training for the 2007 run in London.

But in early March 2007, just at the end of a 13 mile training run, I was all of a hundred yards from home and just slowing down, walking even, when I suddenly pulled up sharply. I thought I'd pulled a muscle in my left calf. I limped home, wondering what I'd done. I'd run a good time for training, maybe 1h 55m or so, and hadn't done anything that I'd expect to cause injury - no tumbles, no uneven ground, no fast pace work.

Recovery Strategy 1 - Ice and Hope

A few days of ice and elevation went by, and it got no better. I had a pronounced limp and really felt the pain going up or down stairs. I upped my dose of Glucosamine and Chondroitin (there's no evidence that they help in injury situations, but I figured it was worth a go)

Recovery Strategy 2 - Ultrasound and Hope

I went to the physio to have it checked out. Her immediate view - not even 10 minutes into the session - was that I'd torn the meniscus cartilage, but she couldn't be sure. Maybe it was just a strain; the optimist in me hoped so. Some ultrasound, some galvani frog treatment (where they strap some wires to your leg and feed a current throught it resulting in lots of rhythmic pulsing of muscles).

Skiing?

I was due to go skiing a week later. The physio sourced a knee brace for me that had near-solid metal bars down either side of the upper quad and calf and a hinge joint at the knee, all aimed at keeping the knee from twisting (and doing further damage)There was no question, of course, of me not going skiing - I'd been looking forward to it for ages and the snow was rumoured to be great. The week of skiing passed without further injury, although any turns where I needed to put weight on my left knee suffered - and the one time I went down a mogul run (entirely by accident - I try and avoid the things normally), the results were calamitous. Even this year, I didn't try that run again.

Recovery Strategy 3 - More Ultrasound, Less Hope

After the skiing, there was no improvement (of course - neither time nor the exertion seemed to have made a difference although at least it wasn't worse). I looked for a different physio to see what someone new would think. A few sessions there, lots of exercises to work my quadriceps and hamstrings, the wearing of various leg braces and what not, and there was no change. Still painful at every step with worse pain on stairs. I'd started to cross roads at crossings, because I couldn't take the risk of needing to dart through a gap between traffic (I was confident I wouldn't make it)

Recovery Strategy 4 - Open It Up

In June (so 3 months after the original injury) I went to see a surgeon. He pushed my left leg this way and that way and within a few minutes said that he was certain I had a tear and that I'd probably had it since the run, not because of anything I'd done since. He didn't need to do an MRI - a waste of time he said given that I'd tried physio for several months. Cartilage tears sometimes happen because of an impact - a forceful twisting of the kind rugby players or football players might experience in a heavy tackle - and they sometimes just happen because bits of the body wear out. Maybe it's running, maybe it's the way that you walk, maybe it's a problem in the alignment of your joints, or maybe it was just my time.

A few days later, in June, I had an operation. It's a simple surgery. You're in and out the same day and can go home, just taking care not to put too much weight on the damaged leg. Within a few days you can cycle and do light weight bearing exercise (to strengthen the quads) and, in theory, after 6 weeks or so you can start running.


Knee Pre Op.jpg

The photo on the top right, taken by the surgeon, is where I think the tear was. The bottom left shot shows the surgeon using the latest hi-tec instrument (it appears to be a downsized boat hook) to sort the cartilage out.

I'm using a demo version of software called "PixelMator" to edit these shots - so there's a logo in the middle of the shot. Once I've played with the software some more I'll register it and replace the shots with pictures without the logo.

Recovery Strategy 5 - Take It Gently After The Op and Then Build The Strength Back

I keep pretty good notes of my training and, looking back at the few weeks after the operation, I can see that I waited a few days and then started cycling (in the gym). The op was on a Friday and on the Monday there was no trace of a limp and only minor pain when walking. Within 10 days I was able to cycle a good 10k, although the pain was still there it was pretty subtle. I felt good

By September (again, 3 months after the operation), I'd jogged a little (on the treadmill and outside) and was ready to try a longer run. I went out for a 6k run and completed it in a careful and slow time of 36 minutes (versus a usual time of 28-30 mins). A week later I tried the same run again, at the same pace, but had to pull up after 3km. I could barely walk let alone run another step. I hobbled home.


Knee Post Op.jpg

This is the post-op photo of the knee. Nothing much to see, although if you look carefully, you'll see from the top left that there is less cartilage in the knee than there was before. It's not easy to tell the difference.

This Isn't Working For Me

I left running alone, went back to the physio and started building up my knee strength again. Maybe my quads were still out of balance.

Six months since the operation, ie January 2008, there was no change. I went back to the surgeon and explained the problem. The next day I had an MRI. It looked liked I'd torn the meniscus again, but in a different place. Maybe it was just weak after the operation, maybe I pushed it too hard or maybe I was just unlucky. Apparently these things happen in perhaps 1 in 5 cases - there's a second tear or a further tear that wasn't quite fixed the first time round.

The MRI - How It Should Look

knee1.tiff

Here's a picture of my knee. I've marked what a normal meniscus cartilage looks like with the red circles. You'll see they're on either side of the knee (in this picture, the front of the knee is on the left). What you should see with a meniscus is a roughly triangular black shape. When your cartilage is torn, you see something like this:


knee3.tiff

This time I've used a green circle to highlight the problem. You'll see there's a white line right through the middle of the previously clear black cartilage cross-section. This is, most likely, a tear.

A Second Operation

At the end of January 2008, I went for a second operation. Same process as last time, in and out within a day. Same plan for recovery, although I took it much easier this time and left it a couple of weeks before I went near an exercise bike.

Recovery Strategy 6 - What More Can I Do?

There's always Rest and Ice, and there's ultrasound, and there's strength building exercises for the quads and hamstrings. I'd been doing all of these, and I carried on doing them. They're the core of any recovery strategy. But I added some new things.

Powerplate exercise. The powerplate is a vibrating platform that supposedly improves recovery from injury, tones and exercises your muscles and so on. Lots of gyms have them now. I've been working on that 5-10 minutes at a time, 4-5 times a week. Does it help? Hard to say, but my weight training poundages are up and I can cycle further in the same time than I used to be able to before the powerplate.

New pills. I've scoured the Internet for any kind of over-the-counter pills that might improve things. I've landed on two that seem to make a real difference. It could all be the placebo effect but, you know what, if they appear to make it better, I'm ok to keep taking them.

The tablets are Super-Cissus RX (only available in the USA but appears not to be a substance banned by the IOC) and Arthrolactin. Both are readily available over the web and won't break the bank.

KneedIt - the last part of the recovery strategy is a new kind of knee wrap called the "KneedIt". It looks really weird (this is definitely not my knee by the way):

200804221405.jpg

But it seems to help. Over the last 4 weeks I've got back into running. I started out at ten minutes and have built up to about 30 minutes so far, running every 3 or 4 days. The distances are nothing to shout about - the furthest I've run so far is 4.25km - but they're a big improvement on being afraid to cross the road in case my knee gave up on me half way across.

So all of these strategies seem to have played a part in getting me back where I am, wherever that is. For now, I'm sticking to wearing the kneedit most of the time, taking the Super-Cissus and Arthrolactin pills every day and hoping that the powerplate is really doing something for me, other than shaking loose every bone in my body.

Is It Fixed?

Probably not. I still have pain when I twist the knee, or when I stand up after a period sitting down. Sometimes it hurts going up or down stairs. I'm hoping that this is all part of the recovery process and that before long that pain will be gone too. I have few options left open to me now, other than trying out cortisone injections. I know that I'm nowhere near needing a new knee, so I'm not going to even think about that for another 20 years at least.

It's been a tough year. When you're used to running 3 or 4 times a week, suddenly not running at all is a big hit - and cycling or working out in the gym don't have quite the same buzz for me. Dealing with the pain with every step, and worse going up and down stairs has been tough too. But it could be worse: I am otherwise fit and mobile and there are plenty of people in worse places than me, so I'm not complaining. I'll keep plugging away and hope that it gets better soon.

Wednesday, April 16, 2008

Marathon Ballots

As if we needed any more evidence that the online world reshapes everything, how about this screen shot from the London Marathon site:

200804161104.jpg

I think I liked the old way where you had a few months to figure out whether you were up for it, get some training in over the summer, fill in the forms (that were liberally dispersed around sports shops throughout the UK) and then post it. Some time in December you'd hear if you had a place. Then you'd get your winter warmer running gear on and pound the streets.

With this new process, they really have no excuse for not telling you in the next 5 days whether you have a place, giving you a whole year to get yourself in gear for running it. Of course, for those who delayed a few days or who haven't got as far in their do list as planning for next year, this will come as a disappointment.

The message is clear though - if you have even half an idea of running the London Marathon 2009, you'd better be online at exactly 9am on 21st April, else you're going to have to work your charity contacts.

Sunday, April 13, 2008

London Marathon 2008

The top three finished inside the London course record, making for 3 of the fastest marathons ever. 1100 people finished inside 3 hours; I'm guessing there were a lot of personal bests run today. For those running, the weather was probably just right: a little cold, some moderate rain and almost no wind. For those at the finish line, me included, there was a little too much rain.

Martin Lel turns the final corner and opens up into an incredible sprint to the finish:

lelopensup.jpg lel1.jpg

Finishers 3 and 4.

goumri1.jpgmergia1.jpg

First Brit home, Dan Robinson (2h 13m), in the right hand picture

wanjiru1.jpg firstbrithome1.jpg

Liz Yelling coming home in a great time (2h 28m) ... and another Brit, Lucy Hasell (2h 40m) clearly feeling the pain

nearlythereyelling1.jpg nearlythereyelling2.jpg aintnopain.jpg

Some more early finishers ... nice tattoos on the right (plus looking very relaxed, compared with the folks on the left)

sprintforit.jpg nicearms.jpg

Geta Wami, second in New York last year, 3rd this time in 2h 25m; James Cracknell going sub 3, no matter the effort required

whenitrains.jpg cracknellsub3.jpg

I thought that it was the end for this lady as she collapsed at my feet. Some quick hamstring massage and she was on her way again, finishing in less than 3h 30m, despite the fall.

dontstopnow.jpg uhohnotthisway.jpg

Three wheels good; First Elvis home, sub 3 hours

3wheelsgood.jpg firstelvis.jpg

Who needs wheels when you have arms that can carry you over the line. This guy stopped 20 yards short, pushed his wheelchair over the line and walked on his hands across the line in under 2 hours,

whoneedsthemwheels.jpg

What the marathon is all about ... got to reach the finish, no matter what. Feels Like Team Spirit as the song might have gone.

youknowithurts.jpg feelsliketeamspirit.jpg

Martin Lel, powering to the finish in a new course record. My favourite picture of the day.

intheairtonight.jpg

Fantastic event, great organisation. I'm hoping to be there next year. Maybe.

Sunday, April 06, 2008

Cold Running

aprilsnow.jpgFor those who looked out of their windows this morning, ready for perhaps the last long run before next Sunday's London Marathon, you have my sympathy.

Last year's run was a challenge for almost everyone because of an unexpected heatwave - I was hot just sitting in the stands watching the finish; I can't imagine what it would have been like running for a few hours in weather like we had this morning, had the marathon been today.

It's not like you're expecting to sign up for the North Pole run when you sign up for London.

The forecast for the end of the week and into next weekend looks much better though - sunny and perhaps 11-13C. Let's see if the forecast holds.