Philip Kendall

What should happen to teams that don't try?

Wed, 01 Aug 2012 22:17:39 GMT

One Olympics-related story which is drifting slightly under the radar is the disqualification of eight women's badminton players for "not trying to win". To cut a slightly long story short, the players knew that a loss in round-robin play would give them an easier match in the knock-out stages, so they actively tried to lose.

Rolf asked the question (behind the Facebook firewall) as to why they were disqualified. As I see it, there are two forces at work here:

It's the job of the players/coaches to do whatever they think is best in accordance with the rules to achieve their goals. In this case, the badminton players took a rational decision to sacrifice the goal of "win this match" as they considered the goal of "win the gold medal" to be more important.
However, it's the job of whoever runs the sport (in this case, the Badminton World Federation) to ensure that the sport is something that people want to watch - and the spectators made it pretty clear that they don't want to watch people trying to lose.

I think the important point here is "in accordance with the rules": if the BWF has an explicit rule that states something along the lines of "you must always try to best to win every match", then the players broke the rule and could reasonably expect some form of sanction because of that. I couldn't instantly see such a rule on the BWF website, but I didn't look very hard.

If the BWF doesn't have an explicit rule along those lines, it all becomes a bit more difficult and under that good-old standby of "bringing the sport into disrepute". However, I think that anyone trying this kind of strategy would be well aware that they were at the very least bending the rules, and sometimes when you bend the rules, you find out that you bent them too far.

Accessing your voicemail

Wed, 02 Feb 2011 21:40:57 GMT

This week's Private Eye claims (in one of its "Number Crunching" sidebars): "1 in 248,832: probability of accidentally tapping in correct sequence of keys to access voicemails". A little bit of thought reveals that 248,832 is in fact 12^5 - the only way I can see this makes sense is if people have 5 "digit" voicemail PINs, in which "*" and "#" are also allowed. Does anyone actually have this kind of PIN for their voicemail, and if not, am I missing something obvious, or has Private Eye just got it wrong?

Asymmetric sports

Sat, 08 Jan 2011 12:49:23 GMT

Most board games (eg chess) are essentially symmetric, in that all players in the game are constrained by the same rules and have the same goal. While there may be some minor differences between players (eg the white player in chess moves first), these are either minor, averaged out by playing multiple games while swapping roles, or both. On the other hand, there are some board games which are explicitly asymmetric: Halatafl (Fox and Geese) is a older example, while the phenomenon is also seen in more modern games such as Scotland Yard, Space Hulk and Last Night on Earth.

Now, while I can think of asymmetric board games, I can't think of any fundamentally asymmetric sports: the closest I can come up with is some things like American football and real tennis where the rules are temporarily different for one player/team than the other, but nothing were the rules are permanently different for one team. Am I missing anything?

Dungeons & Dragons novels

Wed, 29 Dec 2010 16:44:49 GMT

If anybody reading this has any desire for any of the following Dungeons & Dragons novels and is willing to either pick them up or pay postage, let me know. Preference will be given to people I know, people picking things up and people taking away more than one item in approximately that order. Quality varies a bit, but all have sun-bleached spines and none would be described as pristine.

Dragonlance Elven Nations Trilogy
- Firstborn

Dragonlance Heroes
- The Legend of Huma (poor condition)
- Stormblade
- Weasel's Luck

Dragonlance Heroes II
- Kaz the Minotaur
- The Gates of Thorbardin
- Galen Beknighted

Dragonlance Meetings Sextet
- Kindred Spirits

Dragonlance Preludes
- Darkness and Light
- Kendermore
- Brothers Majere

Dragonlance Preludes II
- Riverwind the Plainsman
- Flint the King
- Tanis the Shadow Years

Forgotten Realms Cleric Quintet
- Canticle
- In Sylvan Shadows

Cricket and corruption

Tue, 19 Oct 2010 18:53:24 GMT

The noble sport of cricket this summer was once again in the headlines for all the wrong reasons as three Pakistani bowlers were accused of spot-fixing. One proposal now being floated by the ICC is to use undercover agents to see if players are reporting approaches from bookmakers. What interests me here isn't so much that the proposal has been floated, but that it's been publicly announced. Naively, this would seem to me that it would reduce the prospect of this catching any cheats as they'll know there's a non-trivial chance it's a sting... on the other hand, it may actually be more effective at cutting down on any illegal activity.

I wonder how many actual agents they'll use, or if this is actually a bit of a bluff and the threat of "bookmakers" being ICC agents will be enough to have a significant effect without getting into any possibly ethically dubious areas.

Finance software

Sun, 19 Sep 2010 10:24:13 GMT

Dear lazyweb,

I'm vaguely looking for some software which will let me get a bit of a better handle on what I spend my money on every month. Requirements are hopefully not too stringent: can import data from whatever my bank/credit card providers can export (they'll all seem to manage tab/comma separated value, some of them may do more complicated formats), allows me to automatically classify transactions into a number of "buckets" (entertainment, food, salary, savings, holiday, etc), and lets me get some sort of summary report of those buckets.

About the only other requirement is that it doesn't cost a significant amount of money - I could probably knock up something to do most of what I want in not too much time.

Any suggestions appreciated.

ADE651

Sat, 23 Jan 2010 11:46:34 GMT

It's now looking very much like the scam that is/was ADE651 is going to come to an end pretty quickly, which is unquestionably a good thing (and due credit to Private Eye for having been after this one for a while now). What's surprising me the most isn't that somebody attempted to make a quick buck out of the security situation in Iraq, it's that the Iraqi government has spent £80 million on these things, which are so obviously bogus to anyone with even the most basic understanding of science ("powered solely by the user's static electricity", "detect explosives, banknotes and ivory 1km underground" (paraphrased slightly), "works on nuclear quadrupole resonance or nuclear magnetic resonance", "the theory behind dowsing and the theory behind how we actually detect explosives is very similar"; I could go on). Who authorised the purchase of these things, and why are they not criminally negligent? Am I being too cynical if I expect that there will have been a significant "donation" from ATSC (the manufacturer of the ADE651) to someone in the Iraqi government?

Haiti

Thu, 21 Jan 2010 11:04:30 GMT

Donating in aid of Haiti just because you're going to get free stuff is the wrong reason for donating. But... if you were going to donate something anyway, and you're an RPG fan, you could do lots worse than DriveThruRPG's offer.

About the only thing to note is that DriveThruRPG are requesting you don't actually download everything now as otherwise their servers fall over. Also note there's no time limit on the downloads.

An "interesting" photo

Wed, 25 Nov 2009 14:17:37 GMT

One thing which is obvious to anyone that works in the tech industry (and probably most people that don't as well) is that it's not well populated with women. However, anyone taking a casual look at the team photo (three-quarters of the way down the page) for Assassin's Creed II will notice the large proportion of women there. On the other hand, anyone taking a closer look will note that the entirety of the crowd beyond about 4 rows back is exclusively male...

Colts v Patriots

Tue, 17 Nov 2009 14:27:03 GMT

Despite its appearance, this post isn't really about American football.

On Sunday night, the Indianapolis Colts played the New England Patriots in the NFL. It was a close game, and the result was apparently significantly influenced by one decision from Bill Belichick, the Patriots head coach. Towards the end of the game, he had one of two choices: a low risk tactic or a high risk one (I suspect people either know what those choices were or don't care, so I won't go into them here). One thing the statistical community built up around American football have been saying for the past few years now is that teams are too risk averse: ie they would do better if they used higher risk tactics more than they currently do.

What happened in this case was that Belichick went for the high risk tactic, and it failed: the Colts went on to win the game. This decision has been pretty much universally panned by the media, despite the fact that the stats community are saying the decision didn't actually make that much difference. Of course, the media (and fan) reaction goes a long way to explaining why NFL coaches are more risk-averse than they "should" be, but I still find in surprising that there's such hostility to a decision which by the best objective measures we have wasn't obviously wrong, especially after Moneyball. But that may be because I'm a stathead.

Free Sun hardware

Tue, 17 Nov 2009 11:14:16 GMT

My work is disposing of:

2 Sun Ultra 5s
3 Sun Netra T1s
1 Sun Netra X1

The Ultra 5s and the T1s are believed to work last time we turned them on (3 years ago), but the X1 didn't. Collect from Toft during office hours.

So, Nick Griffin...

Fri, 23 Oct 2009 12:47:57 GMT

If you haven't heard that that Nick Griffin was on Question Time last night, you're either living under a rock or aren't in the UK (if so, Google will give you some sort of clue what happened). What happened was pretty predictable: it turned into a "bash Nick Griffin" session. This isn't a bad thing in any way, and as is fairly normal for him, Mr Griffin was largely ineffective in dealing with it, and apparently surprised by it. I don't know if this is because he's actually so stupid as to not expect it, or if he's just trying to get as much publicity out of this as he can. To me, it just makes him look like a whinger, but I'm not his target audience.

Thus a question... it was fairly safe for the mainstream parties to appear with Mr Griffin as he almost certainly wasn't going to put on a decent show. Would they have done something different if Mr Griffin were actually a top class debater?

Photo backlog (part three)

Sun, 11 Oct 2009 14:37:41 GMT

Karen went on a hen weekend at the start of May.
Karen and I went to the Algarve at the end of May.

Photo backlog (part two)

Sun, 04 Oct 2009 14:17:04 GMT

Karen and I went to Munich in January.

Photo backlog (part one of many)

Sat, 03 Oct 2009 15:17:54 GMT

Mostly of interest to family...

How not to rip off somebody else's code (part 2)

Tue, 29 Sep 2009 07:27:14 GMT

A quick update on the ~~JSSpeccy~~ZX Gamer front:

Mr Boesch has "decided" to remove ZX Gamer from the App Store "to avoid more malicious incriminations and senseless discussions", saying "Apple itself has nothing to do with this decision". While that may be strictly true, it's a case of jumping before it was pushed, as there was already a complaint from me working its way through Apple and I know at least one of the copyright holders of the games Mr Boesch was including in the package had contacted him as well. He's still trying to deny it's a rip-off though.
Mr Boesch is preparing to release "version 1.2" of his app, which he has claimed is "completely rewritten". Needless to say, I'll be keeping more than a close eye on this when/if it does appear.
Matt Westcott has updated JSSpeccy to give it a nice iPhone/iPod Touch interface and speed boost. No-one sane is going to claim it's really any good, but it's a) better than ZX Gamer, b) legal and c) free.
I'm still working on asserting my GPL rights with regards to ZX Gamer. Not too hopeful of anything happening, but you never know.

How not to rip off somebody else's code

Sun, 27 Sep 2009 22:55:08 GMT

The Spectrum fans amongst you may be aware that there is now a Spectrum emulator, called ZX Gamer by Roger Boesch, available for the iPhone. There's one important thing about this emulator, and that is that it sucks. It doesn't just suck a little bit. In fact, it manages to suck so much that it actually runs slower than a real Spectrum (3.5 MHz) on my 3GS (600 MHz). Everyone else who's had the misfortune to download this app agrees, as it's currently got 13 ratings on the App Store, each of which gives it just one star.

Now, there's basically no way you can manage to write an emulator which runs that slowly, unless you write it in a completely inappropriate language. One completely inappropriate language to write a Spectrum emulator in would be JavaScript. However, Matt Westcott is a bit of a nutter (in an entirely good way) and in fact has done just that, calling it JSSpeccy.

Now this starts to get interesting... the three games distributed with ZX Gamer (Batty, Cyclone and Exolon) are three of the example games Matt put up for the JSSpeccy demo. Also, Mr Boesch contacted Matt to ask him some details about JSSpeccy's Z80 core implementation. What's now particularly interesting is that ZXGamer runs at exactly the same speed as JSSpeccy (which happily runs on the iPhone), taking just over 30 seconds to flip from the controls screen to the high score table on Batty. This could all be coincidence, and Mr Boesch has publicly stated that ZXGamer is not based on JSSpeccy. So that's alright then.

However, I'm well known to be a sucker for punishment, so I paid for and downloaded this sucky software, and after backing up my iPhone had a little snoop around the backup directory. I quickly found a couple of files...

$ cat 6573fa774785883cf41752388d5fa9d4b7d1a482.mdinfo

[ binary junk ]AppDomain-com.rogerboesch.zxgame[ more binary junk ]

For the uninitiated, that says that the file 6573... contains data relating to the application name "zxgame.rogerboesch.com". The .mdinfo file contains the metadata, while the actual file data is in...

$ head 6573fa774785883cf41752388d5fa9d4b7d1a482.mddata

function sign_extend(v) {

return v < 128 ? v : v-256;

}

function z80_do_opcodes()

{

while(tstates < event_next_event ) {

var opcode;

Now, let's take a look at the JSSpeccy source....

function sign_extend(v) {

return v < 128 ? v : v-256;

}

function z80_do_opcodes()

{

while(tstates < event_next_event ) {

var opcode;

Wow! Mr Boesch has managed to completely independently come up with his own Z80 core implementation which looks exactly like JSSpeccy's. What are the chances of that happening?

Disclaimer: JSSpeccy is basically a translation of Fuse's Z80 core (which was written by me) into JavaScript. What this does mean is the JSSpeccy is GPL code, and therefore so it would appear is ZXGamer. Where should we go from here?

So, the England cricket team...

Mon, 10 Aug 2009 10:23:47 GMT

So, as most people have probably heard about, the English cricket team capitulated in sorry fashion at the weekend. The leaves the question as to who's going to play in the final, decisive test match.

Cricinfo are reporting that Andrew Flintoff declared himself fit for Headingley, so I can't see any way he won't be playing at the Oval, unless the injury worsens before the match. Other than that, people who I think are almost certainly safe:

Andrew Strauss, Alistair Cook: best batsmen we've had so far in the series.
Paul Collingwood: there are going to be changes in the middle order, but Collingwood has been the least bad up to now.
Matt Prior: batted well, kept well.
Graeme Swann: only possibility of him going is if England go with an all-pace attack.
James Anderson: better than the other quicks so far and less likely to be dropped than Onions.

So, with those 6 and Flintoff, that leaves two batting and two bowling slots up for grabs. Of the batsmen, one of Bopara and Bell will go for Jonathan Trott. That leaves the question of what happens to the other: the names floating around this morning are Marcus Trescothick and Mark Ramprakash. Trescothink I'd love to see back, but his other problems are well documented. Ramprakash? Who knows. I suspect we'll see Bell at 3 followed by Trott at 4.

Of the bowlers, Broad hasn't been impressive all series, Strauss seemed to have no faith in Graham Onions at Headingley and we saw bad Harmison at Headingley. With Flintoff back, it's two from three. Needing to win the match, I'd pick Harmison and Onions. So... Strauss, Cook, Bell, Trott, Collingwood, Prior, Flintoff, Swann, Harmison, Anderson, Onions.

Go vote!

Thu, 04 Jun 2009 08:14:24 GMT

This is a call to everyone eligible reading this to make sure you use your vote today, in particular in the European elections. Due to the proportional representation system used and the fact they're mobilising their support effectively, there's a non-zero chance that some very nasty people might get some seats if the turnout is low. Doesn't matter who you vote for, so long as it's not them.

On the other hand, I was amused to learn that the flyer for the aforementioned nasty people featured a Polish Spitfire, American construction workers, Italian pensioners and an ex-Scots Guardsman who has since described the nasty people as "scumbags".

5 links for a Friday

Fri, 27 Mar 2009 16:23:25 GMT

Unique Killer-Whale Pod Doomed by Exxon Valdez: title says it all
Cricket: the "nightwatchman" strategy: is it really true that sending a nightwatchman in costs a team 25 runs?
Trusting Hardware: On Trusting Trust for a new generation
Price Drop: Stocks, Homes, Now Triple-Word Scores: aka "be careful before changing the rules of the game".
The $300 Million Button: so obvious with hindsight, but how many sites make the same mistake?

Only two links

Fri, 20 Mar 2009 16:19:33 GMT

Molecular Frameworks, the Building Blocks of All Life: yet another attempt to exploit the Long Tail.
Note to the Administration: The AIG Flap is Because of YOU: see also Bank of Scotland, obviously.

Another 5 links

Wed, 18 Mar 2009 16:12:13 GMT

Home Office clueless over its own anti-child porn measures: won't really come as a surprise that the government can't do IT
The Sun Cloud: cloud computing is going to happen, and this is how Sun are doing it
LGA banned words - full list: wondering how many of these are in the recent South Cambs magazine
BBC botnet 'public interest' defence rubbished by top IT lawyer: ethical? unethical? Worthwhile use of my money? Should the BBC be giving money to known criminals, even in the "public interest"?
Intelligent Mistakes: How to Incorporate Stupidity Into Your AI Code: "The AI needs to be more intelligent in order to appear less intelligent."

5 links

Tue, 17 Mar 2009 11:16:11 GMT

An experiment. This may or may not happen every day, and almost certainly won't happen at weekends.

Universities push for higher fees: is this just a play for more government funding?
Jubilation as march on Pakistani capital forces president to climb down over law chief: I'm not going to pretend to understand Pakistani politics
2010 World Cup Qualifying Chances: England basically certain to qualify?
The Most Interesting Statement I Read Today: do big-money (baseball) teams really pay the same for skills as small-money teams?
Graduation rates of tourney's No. 1 seeds vary: ignoring the obviousness of the headline, shows up that "student athlete" means something very different at different universities.

Spectrum vs Virtual Console

Tue, 20 Jan 2009 09:48:49 GMT

A number of games which were released on the Spectrum have also had a release of some form on the Wii's Virtual Console. Some of these games are better in their Spectrum incarnation, whereas some of them are better in their VC incarnation. The question is which is which? With thanks to WoS and Virtual Console Reviews, we can obtain the "definitive" answer...

Compiling Fuse for the Wii

Sun, 18 Jan 2009 10:15:56 GMT

Thanks mainly to Björn Giesler, Fuse now has a pretty functional Wii port, which is available from WiiBrew. When I compiled this, I couldn't find any particularly simple guide as to how to set things up to compile an Autotools application on the Wii, so I'm posting this in case anyone else needs it.

Install the Homebrew Channel on your Wii. You'll almost certainly need a copy of Twilight Princess to do this.
You should now be able to run homebrew (eg the binary release of Fuse) on your Wii.
Install DevkitPro on your build machine. The toolchain is nicely compartmentalised, so you don't need to this in a chroot/VM or the like (with one minor exception: see below).
You should now be able to compile the examples from DevkitPro and run them on your Wii.
Fuse requires one extra library not shipped with the main DevkitPro distribution, libfat. Get the libogc version and expand this under libogc in your DevkitPro installation.
You now need one minor patch to autotool's config.sub to recognise the Wii toolchain. Apply this patch to whatever config.sub you'll be using (on my Debian installation, that's /usr/share/misc/config.sub. YMMV).
You should now be able to compile Fuse using the configure invocation given in Fuse's README.