 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Car Forum / UK Car Forums / Driving (UK group) / December 2005hazard perception test hints
Hi, I have my theory test on Tuesday, and need lots of help with the perception test. I can consistently get 35/35 on the mock tests on the multiple choice so aren't worried about that, but from cold I've *never* passed a mock perception test. 3 days of staring at blurry low resolution AVI clips is sending my eyes funny. It seems to get 5 in most of these you have to be able to be clairvoyant - for example there's a clip with a cyclist riding in from the left... you can't even see the cyclist even on stillframe until the scoring has hit 3 - he's lost in the background pixellation. There's one of a woman with a stick crossing - during the scoring window the smudge of pixels hasn't yet resolved into anything distinguishable from the actual road. According to the software documentation (the software is called 'pass your driving and theory test' by GSP) these are official theory test clips - so I don't expect any better resolution on the day (except blown up fullscreen, so it'll be worse). I went the other way, clicking when any hazard looked like it might possibly develop, and scored 0 because it thought I was cheating. I've always been crap at arcade games... Any hints? Or should I just cancel and save my fee? A couple of months practice I might get it right.. OTOH I've memorised all the practice clips now (done them all about 20 times.. still get zero on some that I think are just mislabled) so it's not a true test... will need to source a lot more if I'm to get this right... know of a good source of practice clips? > Hi, > [quoted text clipped - 13 lines] > smudge of pixels hasn't yet resolved into anything distinguishable from > the actual road. <snip /> I don't believe I had any issues with the clarity of the video clips. Could it be an oddity with your computer. You might find things much clearer on the actual test itself. Good luck, Sean > Hi, > > I have my theory test on Tuesday, and need lots of help with the > perception test. I know it's a lot of BS, I passed mine about 6 months ago. Never just click once for a hazard, click once as soon as you see the hazard, click again about half a second later and then once more about a second later. That's the strategy I used to get me through it. As for triggering the anti-cheat in the real test - I wouldn't worry, on some clips I ended up clicking over 10 times and didn't trigger it. Just don't click in a pattern (once a second for several seconds, etc) and you should be ok. Good luck > Never just click once for a hazard, click once as soon as you see the > hazard, click again about half a second later and then once more about a > second later. That's the strategy I used to get me through it. It's hard not to click once a second... I tend to fall into patterns with stuff like that. I'll see if I can try the half a second/second thing. I've spend the last day trying to work out a passing strategy (having abandoned 'click when there's a hazard' as a definate failure... you either click far too early & miss it, or far too late as the hazard wasn't visible on the screen until after the window - bearing in mind that some of the scoring windows are less than a second long). The best I've ever scored was based on 'click 2/3 times when anything happens on the screen' but you have to be careful not to trip the anti-cheat stuff on a busy clip. The country road stuff is usually OK with a strategy like that (except when I got a zero for not seeing a tractor that due to it being completely invisible until you're almost on top of it). btw. the video quality is down to the player in the testing program - if I play the clips in media player they're almost watchable.. still blurry but not nearly so bad. I guess I could write my own training app... it'd be an all-nighter probably but TBH anything that gives me a cat in hells chance of passing is worth it. Tony > The best I've ever scored was based on 'click 2/3 times when anything > happens on the screen' but you have to be careful not to trip the > anti-cheat stuff on a busy clip. That technique worked for me, got me a pass on my motorbike test. I made 3 rapid clicks every time I saw something that could possibly be considered to be a hazard, possibly clicking 30 times or more on some clips, never once did I trigger the cheat sensor. It would seem that you are not marked down for clicking something that they don't see as a hazard, and you only trigger the cheat thing if you have a regular pattern of clicks. IMO, the hazard perception test is more a test of your passing technique, than of hazard perception. Good luck! > That technique worked for me, got me a pass on my motorbike test. I made 3 > rapid clicks every time I saw something that could possibly be considered to [quoted text clipped - 3 lines] > don't see as a hazard, and you only trigger the cheat thing if you have a > regular pattern of clicks. I wrote some software which showed me each sample clip, with the countdown on the screen 5,4,3,2,1 and started to work out what they were calling a 'developing hazard'. The one thing that was essential that I should have realized earlier.. it's a setup. There's no point in using real-world evaluation. The clip starts.. it takes maybe a second to evaluate all the possible places they could hide a hazard - the trick is to spot it before they start so you have plenty of time to click at the right moment. If there are schoolchildren on the road, you *know* they're going to be running across it in the next few seconds... if there's a van parked on the side of the road you *know* a kamikaze pedestrian is going to run out in front of it... if you're on a dual carriageway someone will overtake from the left or right - guaranteed. It's just a matter of knowing what to look for. Spent hours learning the patterns and found I could predict the setup with reasonable accuracy after a while. Eventuallly got 57. It would have been 62 but I spotted one hazard so early it triggered the cheat thing (saw the sign for a junction on the left and it was obviously someone going to pull out without looking... unfortunately I started clicking too early and it was about 8 clicks before it actually happened, and it scored me a zero). I was a bit shocked as it's not easy to do a test when you've got a bad cold, and a splitting headache.. Tony > The one thing that was essential that I should have realized earlier.. > it's a setup. There's no point in using real-world evaluation. The [quoted text clipped - 7 lines] > going to run out in front of it... if you're on a dual carriageway > someone will overtake from the left or right - guaranteed. It's the "Casualty" effect - they show somebody starting to drive to work or have their tea, and you *KNOW* they're going to be horribly injured at some point in the next ten minutes... > It's the "Casualty" effect - they show somebody starting to drive to work > or have their tea, and you *KNOW* they're going to be horribly injured at > some point in the next ten minutes... heh. :) Makes me wonder exactly what they're testing though... If there was a human component it would be far more valuable as a test of actual hazard awareness IMO. As it is it's a rather overpriced video game (£20 a game!). My instructor told me afterwards that the ones that he did have a flaw in that the exact same car is used for several of the clips, so you could just watch for that car... Tony >Makes me wonder exactly what they're testing though... As it was explained to me, somewhat ambiguously, it's 'developing hazards' rather than 'potential hazards'. I think this means actual hazards that are starting to take place, not what we normally mean by hazards. So a kid playing ball at the roadside isn't a hazard, but when he runs out ino the road he is. Click on 'potential hazards' and you get too many clicks and fail the test. Bit weird, really.  SignaturePaul Rooney > As it was explained to me, somewhat ambiguously, it's 'developing > hazards' rather than 'potential hazards'. I think this means actual > hazards that are starting to take place, not what we normally mean by > hazards. So a kid playing ball at the roadside isn't a hazard, but > when he runs out ino the road he is. Click on 'potential hazards' and > you get too many clicks and fail the test. Bit weird, really. You cant fail for too many clicks. SimonJ <me@mine.net> wrote on Thu, 8 Dec 2005 02:13:13 +0000 (UTC): >> As it was explained to me, somewhat ambiguously, it's 'developing >> hazards' rather than 'potential hazards'. I think this means actual [quoted text clipped - 4 lines] >> > You cant fail for too many clicks. That is entirely incorrect. If you click too often, the software decides you are cheating, giving you 0 for that clip. SignatureDavid Taylor >>> As it was explained to me, somewhat ambiguously, it's 'developing >>> hazards' rather than 'potential hazards'. I think this means actual [quoted text clipped - 7 lines] > That is entirely incorrect. If you click too often, the software > decides you are cheating, giving you 0 for that clip. The software fails you for clicking in a repeated pattern, not for the amount of times you click. >> As it was explained to me, somewhat ambiguously, it's 'developing >> hazards' rather than 'potential hazards'. I think this means actual [quoted text clipped - 4 lines] >> >You cant fail for too many clicks. You certainly can. SignaturePaul Rooney > As it was explained to me, somewhat ambiguously, it's 'developing > hazards' rather than 'potential hazards'. I think this means actual > hazards that are starting to take place, not what we normally mean by > hazards. So a kid playing ball at the roadside isn't a hazard, but > when he runs out ino the road he is. Click on 'potential hazards' and > you get too many clicks and fail the test. Bit weird, really. Nice idea, but the scoring windows are too early for that (at least on the training cdrom.. I've a suspicion that on the real test they're actually a bit more lenient, but it's unlikely I'll ever find out for sure). eg. children running across a crossing. Click when they actually start to run and you get zero - you have to see the *potential* hazard before it actually happens. That's why looking for the setup works... in the real world you make the distinction between potential hazards (and their likelyhood - eg. drunk runs into road, likely. car in front of you decides to do a handbrake turn at 60mph, unlikely) and keep an eye out for the more likely ones. Stuff will happen - it's part of life - but with a healthy amount of paranoia and some forward thinking you can avoid the worst problems. In the test you know the problem can't be avoided (you don't control the car for a start) you're just looking for them to spring it, and timing the scoring window. Tony >> As it was explained to me, somewhat ambiguously, it's 'developing >> hazards' rather than 'potential hazards'. I think this means actual [quoted text clipped - 23 lines] > the car for a start) you're just looking for them to spring it, and > timing the scoring window. So you're saying that people are clicking too late, not too early? Seems unlikely given that most of the complaints are coming from experienced drivers and bikers who have good hazard awareness but fail the test anyway... > So you're saying that people are clicking too late, not too early? Seems > unlikely given that most of the complaints are coming from experienced > drivers and bikers who have good hazard awareness but fail the test > anyway... An experienced driver will click on the approaching hazard early enough anyway... someone with little experience might wait for the actual hazard to occur and be too late. The windows are a little too early for the 'developing' moniker to make a lot of sense (not in all cases, but definately in many - again, I can only go by the windows in the training videos) . Example: You see a car parked on the right with its bonnet open. Common sense says that there's probably someone there... However in the training video there's a scoring window as you approach the car, which hits zero *just before* the person runs out in front of you.. this person that had shown no inclination of doing so (indeed you couldn't actually see them, just assume they were there)... the 'developing' hazard was only developing invisibly behind the bonnet (presumably the owner deciding to commit suicide due to their car breaking down). OTOH I didn't see anything that sneaky on the actual test... the stuff was all fairly obvious, so it may just be down to poor examples on the CD I used). Remember the windows are very small, so repeated clicking is the correct way to score the maximum points, since you can't easily predict exactly *when* the scoring window starts and ends (without having seen the video before, anyway). The trick then is not to click too much. Tony > So you're saying that people are clicking too late, not too early? Seems > unlikely given that most of the complaints are coming from experienced > drivers and bikers who have good hazard awareness but fail the test > anyway... According to the testers when I took mine, Police advanced drivers tend to fail as well. >> As it was explained to me, somewhat ambiguously, it's 'developing >> hazards' rather than 'potential hazards'. I think this means actual [quoted text clipped - 25 lines] > >Tony But if you click as soon as you see the potential hazard, that is too early. You therefore have to click again if and when it becomes what they call a developing hazard, and run the risk of too many clicks and a zero score. You'd also be clicking potential hazards that don't develop, with the same result. SignaturePaul Rooney > But if you click as soon as you see the potential hazard, that is too > early. You therefore have to click again if and when it becomes what > they call a developing hazard, and run the risk of too many clicks and > a zero score. You'd also be clicking potential hazards that don't > develop, with the same result. When I took mine, I clicked on anything that could possibly develop into a hazard, three quick clicks when I first saw it, another 3 quick clicks when the hazard actually started to happen. at least that way I knew I would be clicking in the hazard window, even if I was clicking many times outside it as well. on some of the clips I probably clicked 20-30 times, and I never got a zero. |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2009 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.