The everyday blog of Richard Bartle.

RSS feeds: v0.91; v1.0 (RDF); v2.0; Atom.

11:14am on Friday, 27th February, 2015:

Plain English


This warning message is on one of the projectors I use that sits on an arm attached to the wall above the screen:

This is one of those cases where localisation seems to have taken into account culture rather than just translation. In Spanish and Italian, you're told it's dangerous to hang [from the arm]; in French, you're not told it's dangerous, you're just told not to hang; in Chinese, you're told it's forbidden. In German, however, you 're merely requested not to hang off the arm.

In English, there's nothing about hanging off it at all, you're just ordered flat out not to grab it. I guess that's because we English are such brutes that we don't need to dangle from an arm to break it, just taking hold of it is enough.

As for the picture of the stick man, it looks as if he's about to club something to death.


5:47pm on Thursday, 26th February, 2015:



This is the view from the Waterfront Building at University Campus Suffolk:

It's rather better than the view from my office, which is of, well, other people's offices.


4:06pm on Wednesday, 25th February, 2015:

Who's Whom


I'm pretty good at deciding when to use "who" and when to use "whom". However, I'm having trouble with a line in a (children's) book I'm writing. Ordinarily, it wouldn't matter if I went with the safely-conversational "who" rather than "whom", but the speaker is someone who is a) Victorian and b) a poet. There's no way he'd get it wrong, so I have to make sure I get it right, too.

OK, so this is the line as I have it written at the moment:

"You are indeed whom you are and whom you once were; though not yet whom you will become, I fancy."

What do you think? Is it formally correct or incorrect?


3:31pm on Tuesday, 24th February, 2015:

15 Hours a Week


It's that time of year when second-year students decide what project they want to do in their third year. These final-year projects are big endeavours, worth three times as much as taught modules; students have to spend 15 hours a week on them. Most are selected from lists or projects proposed by members of staff, but the 25% of students who bothered to attend the lecture of the subject are aware that they can also propose their own project topic. All they need to do is get someone to agree to supervise it.

Today, I was asked if I would supervise the development of a game with the following rules:
1) There are 21 balls on the screen.
2) The player and the computer take it in turns to remove 1-4 balls each, with the player going first.
3) Whoever takes the last ball loses.

Hmm. So that means the player loses, because in its turn the computer takes 5 balls minus the number of balls the player took on their turn. This is not a sophisticated game. Much as I'd like to say "instead of 15 hours a week, it took me 15 minutes in total to write a program to play this", I can't honestly do so because it only took me 10 minutes (in Lua). It would have taken me less if I'd been able to get SciTE to work.

So, I won't be supervising that student. Fortunately, they're not on the games module so I don't have to worry that my game design lectures are even less effective than I think they are.


5:07pm on Monday, 23rd February, 2015:

Positive Feedback


Every few years, universities are obliged to contribute to the Research Excellence Framework, in which they describe how much research they have been doing and at what quality. This is an enormous undertaking, which eats through senior researchers' time like nothing else. The point of it is to find out which universities are doing the best research so they can be given more money than those which aren't doing the best research; put another way, it's the means by which research councils determine to give money to the same universities they've always given money to.

One of the criteria upon which universities are judged is "impact". This is a rating of the wider effect research has had on society. It's dealt with through a number of "impact cases", which demonstrate how individual pieces of work have affected the world in general. There's only a limited number of impact cases allowed per department (3 or 4 for one of our size), so being selected as an impact case is therefore quite prestigious as there's a lot riding on it.

For the REF exercise that concluded at the end of last year, my own research on player types was selected to be an impact case.

It was adjudged by the university to be very strong, because it had glowing letters of support from people such as Damion Schubert, lead designer on Star Wars: The Old Republic, explaining how it's foundational to how MMORPGs are designed. However, it was rated weak by the REF committee. The reason given was that although it was clear that player types is used extensively in MMORPGs, the economic impact of this was not demonstrated in the submission to the REF committee.

Now this isn't really fair. It's not fair, because I do actually have data on this and can demonstrate its economic impact through case studies. If the REF committee wanted more data, all it had to do is ask and then we could submit it. However, it didn't tell us that it wanted this data; it merely marked my impact as low on the basis that we hadn't supplied it.

No no no! This isn't an examination! In an examination, if the student doesn't answer the question, you can't go back to them and tell them to re-answer it: the point is to test their understanding. The REF is an information-gathering exercise: if you want information you don't have, you can simply ask for it — or at the very least we should be able to appeal and supply it. It's more like a research paper submission: in those, reviewers can provisionally accept your work but require you go back and make changes. Why on earth, after such a huge undertaking, would a committee not allow missing evidence to be supplied? They're meant to be judging research, not the ability to divine what will impress the committee members.

I dare say the committee members themselves aren't happy with this, mainly because no-one involved with the REF is happy with it. The rules are decreed from on high, however, so we have to live with them.

They could have got the same ranking by simply surveying a couple of thousand parents and asking them where they sent their children to university, weighted by household income...


4:23pm on Sunday, 22nd February, 2015:

Motor Buying


It's my younger daughter's 21st birthday next month, and she's planning on getting a car. I believe that she and my wife have come to some arrangement over how much she has to pay towards it and how much we're paying towards it, but no-one has mentioned it to me so I guess it's something I'm just supposed to know.

Anyway, because of this we spent half of yesterday driving around Colchester looking at second-hand cars. There seem to be three sources:
1) Private sellers. These are the least expensive but you don't have much choice and you can't sue them for millions if they sell you a dud.
2) Second-hand car dealers. These are small businesses that pay tax and stuff. The prices are higher and the choice isn't much better, but you have a better chance of getting something that will work and of pursuing them through the courts if it doesn't.
3) Main dealerships. These sell new cars as well as second-hand ones of the same make. They have a wide choice, the vehicles are all sound and you know exactly what you're getting: ripped off.

It looks as if the main dealerships are too expensive for what you get. Their cheapest cars start at £4,000, and although they probably have some barter-room in there I doubt they'd drop to the £3,000 ceiling my daughter is operating under. We saw some possible candidates at second-hand dealers, but they tended to have crazy mileage. Also, alarm bells ring when a dealer says that they know a car is sound because they gave it its MOT test personally the day before.

Private sales look to be the way to go if you want a car which hasn't done enough miles to drive round the equator twice. You get the fun of trying to decode newspaper small-ads, too. People ought to be putting their veicles on the market around now so they can buy new-registration vehicles next month, which should keep the price suppressed. However, you don't know what's going to come up for sale so it's a bit hit and miss as to what's available. Also, it's probably wise to get the car looked over by a professional before buying, which adds to the cost. All this is irrelevant anyway if the seller won't pick up the phone when you call them...

There is the Internet, of course. There are plenty of cars on offer there, but they still seem to fit the above categories. Some of the second-hand dealers have vehicles on offer that are rather too good to be true, though; I suspect that they're bait, and that if you show up to look at them they'll have been taken "but while you're here, why not look at these other vehicles I have for sale?".

We didn't find anything suitable, so will have to do this all over again next weekend. Apart from the price and mileage considerations, my daughter has managed to narrow her search down to "not in a colour I don't like" and "not a Ford Ka", which is actually quite stringent given the range of colours she doesn't like.

It was so much easier with my elder daughter. For her 21st, she wanted a flute.


1:56pm on Saturday, 21st February, 2015:

Their Future, Our Hands


I was in York yesterday to interview prospective PhD candidates.

So, the Doctoral Training Centre I'm part of, IGGI ("Intelligent Games, Game Intelligence") comprises groups from the universities of Essex, York and Goldsmiths (part of the University of London). Each year, across IGGI, we have something like 10 PhD studentships to award. These are unusual because not only do they cover students' fees, they also come with a stipend (that is, a salary). This makes them highly sought-after. Last year, we didn't have much time to advertise them so most applicants were local to the universities concerned. This year, we did have time to advertise them...

We had something like 60 applications, of which perhaps 50 were credible. We had to turn down some good international candidates because IGGI-wide we only get one non-EU student per year so the good lost out to the utterly outstanding. Each university individually interviewed the remaining students who had applied to them directly, or who had not been specific in their application but who looked a good fit (hint: if you want to apply to IGGI next year, select a particular university rather than not specify, because paradoxically you'll have a better chance of being interviewed that way).

As a result of the local interviews, each university created a shortlist. At Essex, ours was 9 candidates long: these were people whom we thought were definite PhD material, would fit in with the IGGI and other research groups at Essex, and deserved to make it to the final round of interviews in York (which is why I was there yesterday). The problem was, although we'd been working to a shortlist length of 6-8 (yeah, I know 9 is bigger than 8, but they were good candidates!), the practicalities of doing panel interviews meant we all had to limit our shortlists to 6. We had to decide which of 3 perfectly capable students were, through no real fault of their own, not going to make the cut; York and Goldsmiths had similar painful choices. I hope some re-apply next year, but I wouldn't blame them for holding a grudge and going elsewhere instead.

This left us with 18 candidates at the interviews. Although IGGI is a 3-university doctoral training centre, we don't take their target university into account when allocating students (modulo some financial jiggery-pokery we have to do which requires that no single university can award more non-UK EU studentships than it awards UK studentships in a given year). This means we take each student on their own merits, and it's theoretically possible that one university could have all 6 on its shortlist accepted and another have none. We were looking to take at least 9 students IGGI-wide in total, probably 10, and conceivably 11 if strings could be pulled with university high-ups. Our unenviable task was therefore to rank the applicants relative to each other and then take the top 9 or 10. I say unenviable, because the corollary is that we would not be able to take the bottom 8 or 9, even though in absolute terms they were "top" students merely for having got that far.

Apart from numbers, we had to operate under two other constraints to do with financing non-UK students and with choosing between students who were proposing projects too similar to each other, but as it happened those didn't really come into play on the day. Personality issues did when we were unable to separate students on purely academic issues alone (so if we thought one person would get along better with our existing cohort of students, they'd have an edge of the other). This may sound a little iffy, but actually it's incumbent upon us because we got our funding in part because of the "research community" aspect of IGGI we pitched. Other than that, we only ranked students based on academic-related grounds; we'd have done this anyway, but both positive and negative discrimination are illegal in the UK for age, gender, race, disability, sexuality, religion and probably a slew of other things too, so we were always going to be scrupulous in that regard anyway.

So it was that I found myself on a panel of three, alongside IGGI academics from York and Goldsmiths, interviewing 6 prospective candidates one at a time for 45 minutes each, knowing that anything between 0 and 6 of them would be accepted (and the rest rejected). Two other panels did likewise, covering all 18 students by mid-afternoon.

The panels were constructed to ensure that no candidate was interviewed by someone who was up for supervising them; if possible, no panel member should have previously interviewed the candidates, either. This is because the subjective interviewing was done at individual university level; this final stage was to be more objective. It's easy for an academic to enthuse wildly about a candidate whom they want to supervise and can see doing wondrous things; it's not the supervisor who is getting the grant, though, it's the student. The rankings have to be objective, with as much bias eliminated as possible.

Personally, I felt that this approach worked really well. It negated any intra-university political pressures to favour particular candidates, it gave the students a fair chance to show what they were made of, and it gave we interviewers a good idea of overall student quality. It's very hard to make comparisons between candidates bringing wildly differing things to the table, though. Things that always matter — willingless to learn, intelligence, creativity — were shared by all candidates, otherwise they wouldn't have got this far. Other factors are variable or even opposites between individuals: sometimes flair is good, sometimes bad; sometimes a vague vision is bad, sometimes good; sometimes undergraduate achievement matters, sometimes it doesn't; sometimes boasting is bad, sometimes good; sometimes enthusiasm is good, sometimes bad; sometimes ignorance is bad, sometimes good; sometimes over-scoping is good, sometimes bad. It really depends on the combination present in each individual, and what that individual wants to do for their PhD. Add in factors such as nerves and the waters are muddied further still...

Every candidate we interviewed had faults and foibles; sometimes these worked against them, sometimes for them. They all had skills and ability, too. Ultimately, it comes down to the overall balance. For some candidates, the balance was a better fit with IGGI than for others. The former will be getting letters next week inviting them to take up an IGGI place; the latter will get ones telling them they're not. They may later get an offer if someone drops out (which actually happened last year), but most won't. I think we'll also be explaining (at least in some cases) why they didn't get an offer, so they can do something about it and re-apply next year if they're still looking to do a PhD.

Having seen the ranked list, obviously I know who's getting an offer and who's not, but I can't comment on that yet as the individual candidates haven't themselves been told (plus there are some technical hurdles to leap — filling in forms for the funding bodies, that kind of thing). I can say that although the candidates were ranked as individuals, the share across the three institutions didn't work out "unfair" on any of them, which suggests that we're getting a good spread of quality applicants.

I know that there'll be a bunch of happy people we interviewed by the end of next week. That's great, but it's not what I'll be thinking about. I'll be thinking instead about the ones that missed out, even though they were top notch. They were pretty well victims of circumstance.



10:45pm on Friday, 20th February, 2015:



I think next door's cats may be coming into our garden rather more often than I had previously suspected.

I found another track leading off from that same hole, parallel to the hedge, and a third one near another hole in another part of the garden.

Oh well, they probably keep the rats, foxes and bears away or something.


7:45am on Thursday, 19th February, 2015:



I've started cross-posting QBlog to Facebook. Facebook is very pictures-oriented, so I may not be putting stuff there that doesn't come with an image to attract attention. Then again, maybe I should only cross-post imageless entries, given the advertisement Facebook decided to offer me based on the one image I've uploaded so far:

Rather harsh in its judgement, I thought.


2:35pm on Wednesday, 18th February, 2015:



Here's an extract from a 24-year-old research paper:

[Ian Makeman, Dave Lewis and Jon Crowcroft: Traffic Analysis of Trans-Atlantic Traffic. University College London, 1991.]

Translation: in 1991, at least 11% of the Internet packets crossing the Atlantic belonged to MUDs.

Game worlds aren't the force they used to be...


7:46am on Wednesday, 18th February, 2015:

String Theory


In yesterday's pre-election letter from the Church of England's Bishops to their parishioners, there's a passage (number 16) on accumulation of power. This is what it says:

Christians believe that God alone is Lord of creation and our allegiance to any other system or power is only possible so long as it makes no claims which compromise our allegiance to God. Christians should be wary of accumulations of power wherever they take place. They should be as reluctant to live under an overweening corporate sector as under an overweening state. Where the state or the market, or any other powers, claim too much and stifle human flourishing, people are divided from one another and God's sovereignty is mocked.

I don't get how you can be both "wary of accumulations of power" and yet like the idea that the entire universe was created by a single individual who pulls all the strings.

This is why I'm not a Bishop. Well, this and the fact I'm an atheist.


2:11pm on Tuesday, 17th February, 2015:



Here's the sun rising over the mountains from our house this morning:

Well, it would be if we had mountains. Sadly, those are clouds.


4:09pm on Monday, 16th February, 2015:

Lousy Editing


I recently wrote an article for an academic journal. The general editor wanted me to make some minor changes, one of which was to update a reference.

The reference in question was to Robert Burns' 1786 poem, To a Louse, which includes the lines:
O wad some Pow'r the giftie gie us
To see oursels as ithers see us!

I tried contacting Mr Burns to see if he'd brought out anything more recent on the topic, but I couldn't find his email address.


5:14pm on Sunday, 15th February, 2015:

Bunged Up


I've spent today unable to hear out of my right ear. I've been waking up in the morning for the past couple of weeks with it bunged up, but it's always cleared within a few minutes. Not today. I'm going to have to go to the pharmacist tomorrow to buy the ear canal equivalent of drain unblocker. It's on the same side as my bad eyelid, so I'm missing quite a bit of sensory information on my right.

Looking at it positively, though, at least it means that my wife can't sneak up on me on my right and say something loudly that makes me jump.


12:12pm on Saturday, 14th February, 2015:

Three Men


Here are three things that happened on my trip to Sainsbury's today. I saw all three, but only one happened to me. See if you can guess which one.

The Helpful Shopper
A man is walking to get a trolley. He passes an old man on the way who has a trolley. Clearly, the old man is taking his trolley back to the trolley park. The man, wishing to help, says, "I'll take that!" and grabs hold of it at the side. He walks off, taking the handle as the old man releases it. Only when he reaches the building's entrance and looks back does he notice that the old man has had to get a replacement trolley. He hadn't been returning a trolley after his shop, he'd been going to do his own shop. The man will spend the next half an hour in Sainsbury's attempting not to encounter the old man again.

The Sticky Trolley
A man is pushing his trolley when he notices that it's a bit stiff. Believing that there is something lodged under the front wheel, he moves off apace, hoping to dislodge it. He travels to the end of the aisle this way, but to no avail. He decides to stop and remove whatever is stuck under the wheel manually. Nothing is stuck under it. Someone has put their basket on the floor while looking at products on the shelf and the man has pushed their basket half the length of an aisle. He leaves it where it is and escapes down the next aisle.

The Not Brian
A man arrives at the back of his car with a trolley full of shopping. It's a hatchback. He lifts up the door and stares in confusion. What's all this stuff doing in the back of his car? He looks around. Meanwhile, an old lady inside the car is shouting in a panicked voice, "Brian? Is that you Brian? Brian!". The man's own car is the next one along.

So which one of these happen to me?

Answer, -1 encrypted: sgd lhcckd nmd.


Latest entries.

Archived entries.

About this blog.

Copyright © 2015 Richard Bartle (richard@mud.co.uk).