Re: Ask Graeme?
Posted: Mon Jan 27, 2014 9:46 pm
Many thanks Graeme. For both my questions I'm surprised at just how many occasions those scenarios occurred. I accept your judgement as the final arbiter of what is ironic too.
A group for contestants and lovers of the Channel 4 game show 'Countdown'.
http://www.c4countdown.co.uk/
Treat the episode counts in this answer as approximate. The recap writer will have calculated the max for older games using the ODE2r, so words that would be accepted now might not have been accepted then, and this of course means the recorded maxes might not be right.Jon Stitcher wrote:0 max games....
Obviously the games where players scored 0 were zero max games I just wondered how many contestants scored 0 maxes in a game, did any player win whilst scoring 0 maxes, what's the fewest number of maxes a winner has scored, any games where neither player hit a max?
Interestingly, both players from episode 1614 went on to make the quarter finals.Graeme Cole wrote:[Only two games have seen neither player max a round: episode 1189 and episode 1614.
Equally interestingly, Philip Griffiths seems to have maxed R1 in 1189, and both players maxed R7 in 1614.Gavin Chipper wrote:Interestingly, both players from episode 1614 went on to make the quarter finals.Graeme Cole wrote:[Only two games have seen neither player max a round: episode 1189 and episode 1614.
You just made me doubt myself, and I had to think about it for 5 minutesm but they are different:Graeme Cole wrote:Yes, it is. There doesn't seem to be a six letter word which does not have at least four letters in common with some seven letter word.JackHurst wrote:Could you also do the same for 6 letter words? I imagine the list would be empty.
Isn't this the same thing as a six that's always the max?JackHurst wrote:If you still have time on your hands, could you produce the list of 6 letter words that you cannot add letters to and reorder to get a 7, 8 or 9?
Would CLAMORED have been OK back then for R1 in 1189?Clive Brooker wrote:Equally interestingly, Philip Griffiths seems to have maxed R1 in 1189, and both players maxed R7 in 1614.Gavin Chipper wrote:Interestingly, both players from episode 1614 went on to make the quarter finals.Graeme Cole wrote:[Only two games have seen neither player max a round: episode 1189 and episode 1614.
Yes, I think so. It looks as though I was wise to equivocate slightly.Innis Carson wrote:Would CLAMORED have been OK back then for R1 in 1189?Clive Brooker wrote:Equally interestingly, Philip Griffiths seems to have maxed R1 in 1189, and both players maxed R7 in 1614.
Probably. American spellings were allowed for ages before they were banned.Innis Carson wrote:Would CLAMORED have been OK back then for R1 in 1189?Clive Brooker wrote:Equally interestingly, Philip Griffiths seems to have maxed R1 in 1189, and both players maxed R7 in 1614.
I see what you mean. So a list of sixes for which there is no seven, eight or nine which has six letters in common with that six.JackHurst wrote:Thanks for thet Graeme, that's an interesting list.
You just made me doubt myself, and I had to think about it for 5 minutesm but they are different:Graeme Cole wrote:Yes, it is. There doesn't seem to be a six letter word which does not have at least four letters in common with some seven letter word.JackHurst wrote:Could you also do the same for 6 letter words? I imagine the list would be empty.
Isn't this the same thing as a six that's always the max?JackHurst wrote:If you still have time on your hands, could you produce the list of 6 letter words that you cannot add letters to and reorder to get a 7, 8 or 9?
The list I asked for which you generated and is empty should be a list of sixes which always appear as a max, which is a subset of the list of sixes with no stemming.
I'll think of an example using 7s to illustrate the idea: MOUJIKS appears on the list of always max words, so by definition of that list, MOUJIKS has no stems, so cant be on the no stem list, but SKYGLOW, whilst not on the list of always max words, does appear on the list of words with no stems. So there either by an 8 with 6 letters in common with skyglow.
I somehow doubt it. I can't even think of anyone who has done it with 4. Loads have done it with 3.James Laverty wrote:Have any contestants appeared in episodes hosted by all five presenters? If not who holds the record for appearing with the most presenters
I somehow doubt it. I can't even think of anyone who has done it with 4. Loads have done it with 3.James Laverty wrote:Have any contestants appeared in episodes hosted by all five presenters? If not who holds the record for appearing with the most presenters
Junaid Mubeen, Anita Freeland, Nick Wainwright, Conor Travers, Mark Tournoff, Andrew Hulme, Steven Briers, Kai Laddiman, Jon O'Neill, Chris Wills, Kirk Bevins and David O'Donnell are the ones I can think of who have appeared with three. Of those, only Junaid and Anita didn't appear in the 30BC.James Robinson wrote:I somehow doubt it. I can't even think of anyone who has done it with 4. Loads have done it with 3.James Laverty wrote:Have any contestants appeared in episodes hosted by all five presenters? If not who holds the record for appearing with the most presenters
Yeah I could think of loads with three. Thought Chris Wills had been on with four for some reason but I was mistakenJack Worsley wrote:Junaid Mubeen, Anita Freeland, Nick Wainwright, Conor Travers, Mark Tournoff, Andrew Hulme, Steven Briers, Kai Laddiman, Jon O'Neill, Chris Wills, Kirk Bevins and David O'Donnell are the ones I can think of who have appeared with three. Of those, only Junaid and Anita didn't appear in the 30BC.James Robinson wrote:I somehow doubt it. I can't even think of anyone who has done it with 4. Loads have done it with 3.James Laverty wrote:Have any contestants appeared in episodes hosted by all five presenters? If not who holds the record for appearing with the most presenters
For each game, I've recalculated who would have won based only on the letters rounds. If this results in a tie, the outcome of the original game is used. Then each player's run ignores any games they played after one they would have lost. From that I've worked out the number of wins and points each player has left. As you say, we can't answer the whole question, because although we can say that, for example, Mark Deeks would have lost his first game and Jack Worsley would have lost after three wins, we can't say how many games someone would have won after they actually lost. In other words, this recalculation can turn an octochamp into a zero-time winner, but someone who lost a game can only gain one win. For example, James Robinson would have beaten Patrick McCurdie in his third game, but I've only counted him as having three wins because we don't know how many he would have won afterwards. Also, for players who still have 8 wins only counting letters rounds, they might have faced different opponents so the outcomes of those games might have been different.Charlie Reams wrote:Somewhat vague question, feel free to formalize this however is convenient: if only letters rounds counted, how different would the seedings be in a typical series? I'm imagining counting wins and points for the letters rounds up to and including the first game each contestant would've lost (if any) without numbers and conundrums. (This is a bit artificial because you can discount points following games a contestant would've lost, but you can't say much about the impact of the player who actually lost but would've won. But hey.)
Code: Select all
Series 69
Dylan Taylor 8 613
Jen Steadman 8 597
Glen Webb 8 593
Bradley Cates 8 523
Alex Fish 8 484
Zarte Siempre 5 388
Abdirizak Hirsi 5 386
June Glasspell 5 307
Code: Select all
Series 68
Andy Platt 8 620
Giles Hutchings 8 609
Joe McGonigle 8 417
Jill Hayward 4 220
Chris Ball 3 223
Philip Jarvis 3 204
Peter Fenton 2 164
Stuart Scholes 2 145
Code: Select all
Series 67
Paul James 8 572
David Barnard 8 537
Tia Corkish 8 505
Liam Shaw 7 490
Chris Marshall 5 342
Heather Styles 4 325
Philip Jackson 4 298
Sohail Virdi 4 288
Series 66
Peter Lee 8 572
Jonathan Rawlinson 8 565
Suzi Purcell 8 470
Victoria James 6 402
Mark Murphy 5 324
Alison Shipman 5 323
Rob Gibney 4 315
Ben Nicholson 4 254
Series 65
Graeme Cole 8 568
Paul Keane 8 495
Dave Taylor 8 489
Matt Croy 6 413
Carl Williams 4 321
Nikki Roberts 4 276
Jon Elmer 4 241
Jayne Wisniewski 3 261
Series 64
Adam Gillard 8 612
Edward McCullagh 8 602
Tom Barnes 8 586
Andy McGurn 8 505
Mary Adie 5 367
Andrew McNamara 5 325
Michelle Nevitt 4 309
Mike Pickering 3 217
Series 63
Jack Hurst 8 655
Eoin Monaghan 8 609
Daniel Pati 8 589
Marcus Hares 8 572
Scott Gillies 8 555
Tom Rowell 8 532
Peter Godwin 7 510
Michael Chadwick 7 481
Series 62
Peter Zyss 7 463
Lee Graham 7 458
Danny Pledger 7 448
Claudia Tyson 6 414
Dominic Travers 5 360
Raheel Mirza 5 347
James Rawson 4 338
Kevin Davis 4 312
Series 61
Andrew Hulme 8 646
Chris Davies 8 626
Innis Carson 8 598
Ryan Taylor 8 557
Jeffrey Burgin 8 517
Jacqueline Baker 5 355
Bob De Caux 5 355
Tom Allerton 4 296
Series 60
Kirk Bevins 8 644
Jimmy Gough 8 578
Shane Roberts 8 531
Neil Zussman 8 529
Stephen Porritt 4 234
Julie Russell 3 240
Kate Richardson 3 210
James Robinson 3 195
Series 59
Charlie Reams 8 579
Martin Bishop 8 566
Debbi Flack 6 417
Denis Kaye 3 227
John Matthews 3 221
Tony Gilgun 3 208
Samira Mohamed 3 177
Mike Lambert 2 206
Junaid Mubeen, Andrew Hulme, Chris Wills, Kirk Bevins, Mark Tournoff, Jon O'Neill, Conor Travers, Steven Briers, Anita Freeland, Nick Wainwright, David O'Donnell and Kai Laddiman have all appeared alongside three different presenters, which is the record. (Edit: just realised that Jack already answered this, and completely correctly.)James Laverty wrote:Have any contestants appeared in episodes hosted by all five presenters? If not who holds the record for appearing with the most presenters
Almost none.Jack Worsley wrote:What evidence is there (if any) to suggest that having the letters pick gives you an advantage for the round?
I checked that as well, but the results were mostly the same. In the new 15-rounders up to the end of last series, 480 of the 1,370 letters rounds were won by the picker and 477 were won by the non-picker. Of the 28677 old 15-rounder letters rounds, 7329 were won by the picker and 7241 were won by the non-picker. In the 30th Birthday Championship, the difference is (a little bit) more pronounced: in 440 letters rounds, the picker won 75 and the non-picker won 65.Clive Brooker wrote:I'd assumed that the questioner was asking whether the picker has a better chance of winning the round, not maxing it.
I expect the effect would be quite small. The main advantage I can think of is the case where, for example, you've got a safe seven and a risky eight, and your opponent declares six, so you declare your seven and your eight wasn't valid.Clive Brooker wrote:Intuitively, I'd expect the player declaring second to have the advantage in a letters round, this effect possibly swamping any minuscule advantage gained from being the picker. Is the information there to look at that? I'm not sure whether all the presenters - Des O'Connor in particular - were entirely consistent in their approaches.
It would be valid to brush over this fact as unimportant based on the tiny sample size, but to brush it aside as an insignificant margin (which I get the impression you have) is ridiculous! That gap is by no means insignificant in a game of such fine margins.Graeme Cole wrote:In the 30th Birthday Championship, the difference is (a little bit) more pronounced: in 440 letters rounds, the picker won 75 and the non-picker won 65.
Looking at this, in 140 trials the probability of getting 75 or more when there is a 50% chance each time is over 22%, so it's not particularly statistically significant. Even less so when you do a two-tailed test, which you should be doing anyway because you could argue that not picking the letters gives you more time to find words. So while the ratio 75:65 is quite big if it's statistically significant, it's not in this case.JackHurst wrote:It would be valid to brush over this fact as unimportant based on the tiny sample size, but to brush it aside as an insignificant margin (which I get the impression you have) is ridiculous! That gap is by no means insignificant in a game of such fine margins.Graeme Cole wrote:In the 30th Birthday Championship, the difference is (a little bit) more pronounced: in 440 letters rounds, the picker won 75 and the non-picker won 65.
It did for me in one round in the match I played.Jack Worsley wrote:What evidence is there (if any) to suggest that having the letters pick gives you an advantage for the round?
Great story bro. I wish I had the skillz to pull that off.Martin Thompson wrote:It did for me in one round in the match I played.Jack Worsley wrote:What evidence is there (if any) to suggest that having the letters pick gives you an advantage for the round?
The first 8 letters I picked were U I A F D S O G. I realised that if an E came up I would have the word GAUDIES. I asked for a vowel in the hope that an E would be selected, and that's what happened.
It's not an advantage if your opponent spots it too though. It's the whole reason countdown is better than scrabble. You can be the best scrabble player in the world but if you're getting shitty tiles compared to you're opponent you wont win. Countdown is a level playing field with both players playing off the same 9 letters.Martin Thompson wrote:It did for me in one round in the match I played.Jack Worsley wrote:What evidence is there (if any) to suggest that having the letters pick gives you an advantage for the round?
The first 8 letters I picked were U I A F D S O G. I realised that if an E came up I would have the word GAUDIES. I asked for a vowel in the hope that an E would be selected, and that's what happened.
It's not at all unlikely to not see Q, Z and X in a Scrabble game for two or three games in a row, and anyway, the point is that luck isn't much of a factor in CD, whereas it can greatly affect a game of Scrabble. Whether it evens out is moot.Dave Preece wrote:Scrabble tiles 'luck' evens in no time at all, so that's irrelevant, surely?
And Glen Webb had two nines when he lost to Callum. I would assume there'll be a few more, particularly in the 30th Birthday Championship.Mark James wrote:I beat Abdi and we both got two nines.
There was also Ross Mackenzie, who got TOTALISES and HERALDING, but Chris Davies got those in his 139.Bradley Cates wrote:And Glen Webb had two nines when he lost to Callum. I would assume there'll be a few more, particularly in the 30th Birthday Championship.Mark James wrote:I beat Abdi and we both got two nines.
What a coincidence. So did Ross win his game?James Robinson wrote:There was also Ross Mackenzie, who got TOTALISES and HERALDING, but Chris Davies got those in his 139.Bradley Cates wrote:And Glen Webb had two nines when he lost to Callum. I would assume there'll be a few more, particularly in the 30th Birthday Championship.Mark James wrote:I beat Abdi and we both got two nines.
Jon Corby wrote:So basically it happens pretty much every single game.
He doesn't mean single games as in people playing on their own, because it would be impossible in that case, if that was your concern.Johnny Canuck wrote:Jon Corby wrote:So basically it happens pretty much every single game.
I'll have to pass on this for now, as I've just settled down to answer this question and realised that my database doesn't record which words DC got.Gavin Chipper wrote:What's the average number of maxes per game that the studio gets collectively? Any words or numbers declarations outside the time count, and also it can be said that the studio got the conundrum. So basically what's 15 minus the number of rounds with beaters available?
Nine players including Jack Hurst, up to the end of series 69. The full list is Terry Rattle, Matthew Shore, Micheal Harris, Ross Mackenzie, Lesley Hines, Chris Davies, Jack Hurst, Abdirizak Hirsi and Glen Webb.Jon Stitcher wrote:When Mark beat Bobby last week, Bobby got two nines and still lost.
Obviously we all remember Jack Hurst losing with two nines in the 30th birthday final.
Have any other players lost a game where they scored two nine letter words?
Was PREFUCK really valid back then?Graeme Cole wrote:Lesley Hines
Since Series 64, players are sorted by number of wins then by number of points. Points from both wins and losses count. Points from sudden death conundrums aren't counted, and players with the same number of wins and points are sorted by the number of conundrums correctly guessed, then highest individual score.Tom wrote:In seedings, are these still decided by number of wins then number of points?
If 2 players have the same aggregate score and number of wins, would the higher finisher still be the one who has the higher individual score in a heat game and if both players were then still matched, would it then boil down to 2nd highest, 3rd highest individual score etc to determine who finishes higher?
Under the 15 round format, does anyone know what the closest points difference has been between 8th and 9th seed? I've always thought the idea of having an 8 vs 9 play off sounds nice but isn't really fair for the 8th seed if they have a higher points total.
Whilst the chances of it happening are unlikely, if 2 players had the same number of wins and aggregate score plus the same highest individual score and the 8th seeding place in the finals was at stake, I think it would be brilliant to see a play off/special game on the final heat game of the series to determine the 8th seed and the winner plays straight afterwards in the 1st QF.
Yes.Tom wrote:In seedings, are these still decided by number of wins then number of points?
Countdown Team would know the answer to this better than me, but last time I enquired about it (which was 2011), it was that if two players are level on wins and points, the player with the most conundrums solved gets the higher spot. After that I presume it's whoever got the highest score in an individual game, and I don't know how they do it after that.Tom wrote:If 2 players have the same aggregate score and number of wins, would the higher finisher still be the one who has the higher individual score in a heat game and if both players were then still matched, would it then boil down to 2nd highest, 3rd highest individual score etc to determine who finishes higher?
The rules that determine the seedings have changed over the years, but as far as I can tell there were a few cases where the 8th and 9th seeds were particuarly close.Tom wrote:Under the 15 round format, does anyone know what the closest points difference has been between 8th and 9th seed? I've always thought the idea of having an 8 vs 9 play off sounds nice but isn't really fair for the 8th seed if they have a higher points total.
I believe there is (or at least was) a whole list of criteria to separate contestants who have the same number of wins and points, so they wouldn't have to resort to a play-off. I have a dim memory of Richard Whiteley talking in detail about the seed order rules in one show, and I seem to remember that "fastest conundrum time" was one of the things used to break a tie if other things fail. This was about 500 years ago though.Tom wrote:Whilst the chances of it happening are unlikely, if 2 players had the same number of wins and aggregate score plus the same highest individual score and the 8th seeding place in the finals was at stake, I think it would be brilliant to see a play off/special game on the final heat game of the series to determine the 8th seed and the winner plays straight afterwards in the 1st QF.
The "since series 64" bit appears to be incorrect - the decider after wins and points has been number of conundrum solves since at least series 60, when Cate Henderson and Jimmy Gough were separated in this way (also see this post).
Words sorted by the number of times they have been validly offered up to the end of series 69...Gavin Chipper wrote:I don't remember this being asked - what are the most commonly offered words ever on Countdown? Like top 50 or something.
Code: Select all
1. RATION 126
2. TRAINED 125
3. TRAILED 107
4. RATIONS 102
5. ROASTED 96
6. POINTED 92
7. PAINTED 90
8. LOITERS 78
9. PANTIES 77
10. COATED 76
11. COASTED 68
12. DONATES 67
13. GLOATED 66
13. SOLDIER 66
13. STAGED 66
16. FLOATED 63
16. ORANGES 63
16. WAITER 63
19. PAINTER 62
19. RELATION 62
21. BOASTED 61
21. MOISTER 61
21. REASON 61
24. STONED 60
25. FLOATER 58
25. GORIEST 58
25. STRAINED 58
28. GOITRE 57
28. LOITER 57
28. PRAISED 57
28. RADIOS 57
32. POINTER 56
33. DREAMS 55
33. POSTAGE 55
35. ELATION 54
35. MOANERS 54
35. PLAITED 54
35. POLITE 54
39. COASTER 53
39. LOANERS 53
39. MOIST 53
39. RATIOS 53
43. IMAGES 52
44. FASTEN 51
44. STAINED 51
46. FAINTED 50
46. TOILED 50
46. WAITED 50
49. POUTED 49
50. FOISTED 48
50. PARTIED 48
50. WAITERS 48
Nope, the thing used between Series 39 and 63 applies to 60; Gough had a score of 116, Henderson had 113.Graeme Cole wrote:The "since series 64" bit appears to be incorrect - the decider after wins and points has been number of conundrum solves since at least series 60, when Cate Henderson and Jimmy Gough were separated in this way (also see this post).
So why was Cate put above Jimmy then?Rhys Benjamin wrote:Nope, the thing used between Series 39 and 63 applies to 60; Gough had a score of 116, Henderson had 113.Graeme Cole wrote:The "since series 64" bit appears to be incorrect - the decider after wins and points has been number of conundrum solves since at least series 60, when Cate Henderson and Jimmy Gough were separated in this way (also see this post).
Thanks for that, Graeme. No LEOTARD then!Graeme Cole wrote:Words sorted by the number of times they have been validly offered up to the end of series 69...Gavin Chipper wrote:I don't remember this being asked - what are the most commonly offered words ever on Countdown? Like top 50 or something.
Code: Select all
1. RATION 126 2. TRAINED 125 3. TRAILED 107 4. RATIONS 102 5. ROASTED 96 6. POINTED 92 7. PAINTED 90 8. LOITERS 78 9. PANTIES 77 10. COATED 76 11. COASTED 68 12. DONATES 67 13. GLOATED 66 13. SOLDIER 66 13. STAGED 66 16. FLOATED 63 16. ORANGES 63 16. WAITER 63 19. PAINTER 62 19. RELATION 62 21. BOASTED 61 21. MOISTER 61 21. REASON 61 24. STONED 60 25. FLOATER 58 25. GORIEST 58 25. STRAINED 58 28. GOITRE 57 28. LOITER 57 28. PRAISED 57 28. RADIOS 57 32. POINTER 56 33. DREAMS 55 33. POSTAGE 55 35. ELATION 54 35. MOANERS 54 35. PLAITED 54 35. POLITE 54 39. COASTER 53 39. LOANERS 53 39. MOIST 53 39. RATIOS 53 43. IMAGES 52 44. FASTEN 51 44. STAINED 51 46. FAINTED 50 46. TOILED 50 46. WAITED 50 49. POUTED 49 50. FOISTED 48 50. PARTIED 48 50. WAITERS 48
Going on from this point, how often have finalists qualified for the finals but not taken part? Can remember it happening for Rachel Moran in S68 and who ever it was for Michelle to sneak in S64 but I'm guessing it's happened a few more times over the yearsIn Series 63, Michael Chadwick beat Peter Godwin to the #8 spot by one point. However, it turned out that Michael Chadwick didn't turn up for the finals, even though he was practising on apterous just a couple of days before the finals were filmed. I seem to remember he had to fly off to Singapore at short notice. So Peter Godwin was promoted to #8 seed.
Mark Deeks wrote:Was PREFUCK really valid back then?Graeme Cole wrote:Lesley Hines
I don't have an ODE2r, so I can only assume so.Mark Deeks wrote:Mark Deeks wrote:Was PREFUCK really valid back then?Graeme Cole wrote:Lesley Hines