Pseudo-objectively, Who Has Had The Best Octo-run?

[David O'Donnell]

While gripped by another bout of seemingly terminal insomnia I have been pondering this question. I decided to see if there was a fair way you can objectively measure the performance of different contestants. I have used a par system like the gameshow guy does on his website and as Jim Bentley did a while back. I have tried to incorporate many aspects of measuring a performance and tried to weight them accordingly. I don't think I have got the weighting right but thought I'd share with you anyway.

Letters

Maxes score -1, you get an additional -1 for a Darren.
One less than a max is par, while two less than a max is +1. Worse than this is +2 and the same goes for having a disallowed word.
As nines are scored differently I felt a player should really be rewarded for getting a nine and punished if they miss one so I gave it +7/-7 either way, there is no max or Darren point added. In order to compensate for a player not encountering as many nines as another I added whatever the actual number of available nines. This does not take into account how difficult the spots were and I am sure it may be possible to have a fairly objective definition of a difficult nine and scale accordingly (likewise for conundrums or even spots in letter rounds and number rounds generally). However, not even I was that bored.

Numbers
Getting the numbers while the opponent doesn't is -4, if the opponent gets it too then it's -3. Within one of the target is -2, within 5 is -1. Within 10 is par and if worse than that or your opponent beats you in the numbers it's +4.

Conundrums
I wanted to reward speed here so sub 2 sec spots are -4, sub 5 sec spots -3 and all other spots are -1 (probably a bit harsh). A missed conundrum is +4.

I restricted this to contestants getting 850+ and to the eight games (if you hadn't guessed from the thread title).

Have had to edit this because it doesn't make much sense, still not ideal (have tried formatting but am useless).

Letters (ranked in order):
Julian Fell -93
Craig Beevers -92
Chris Davies -87
Innis Carson -74
Kirk Bevins -67
Stewart Holden -66
Matthew Shore -59
Andrew Hulme -58
Conor Travers -56
David O'Donnell -49
Tom Hargreaves -40
Jon Corby -28
Chris Cummins -9

Numbers:
Chris Cummins -77
David O'Donnell -74
Craig Beevers -71
Kirk Bevins -68
Tom Hargreaves -66
Andrew Hulme -63
Conor Travers -62
Jon Corby -60
Chris Davies -54
Innis Carson -51
Julian Fell -49
Matthew Shore -47
Stewart Holden -30

Conundrums:
Julian Fell -24
Chris Davies -23
Innis Carson -23
Andrew Hulme -21
Stewart Holden -20
Craig Beevers -18
Conor Travers -16
Chris Cummins -14
Tom Hargreaves -13
Kirk Bevins -12
David O'Donnell -7
Jon Corby -4
Matthew Shore 4

Totals:
Craig Beevers -181
Julian Fell -166
Chris Davies -164
Innis Carson -148
Kirk Bevins -147
Andrew Hulme -142
Conor Travers -134
David O'Donnell -130
Tom Hargreaves -119
Stewart Holden -116
Matthew Shore -102
Chris Cummins -100
Jon Corby -92

Average maxes per game:

Julian Fell 11.875
Craig Beevers 11.75
Kirk Bevins 10.625
Stewart Holden 10.625
Chris Davies 10.5
Conor Travers 10.5
David O'Donnell 10.25
Innis Carson 9.5
Andrew Hulme 9.5
Matthew Shore 9.5
Jon Corby 9.125
Chris Cummins 9
Tom Hargreaves 8.25

Craig had the most Darrens (22), and four players (Craig ? [need to check this thought he did], me, Chris C, and Tom H) did not concede points on the numbers.

I know there are numerous flaws with my approach but I like how, whatever way you look at it, Julian Fell's performance has stood up. Oh yeah, I probably miscounted a bit so there is a margin for error here!

[Gavin Chipper]

I know there are numerous flaws with my approach but I like how, whatever way you look at it, Julian Fell's performance has stood up. Oh yeah, I probably miscounted a bit so there is a margin for error here!

Obviously no system is perfect but if people come up with their own systems and they all end up with the same one or two people at the top, then the pecking order gains credibility.

It's no surprise to me to see Craig a the top and Julian second, although Kirk's position surprised me a bit.

[David O'Donnell]

Yeah, surprised me too! It proved to me that the system I used must be wrong!

[JimBentley]

I reckon David's system as outlined above is as good as any other; my run scores +2 on letters, -62 on numbers and +3 on conundrums for a -57 total, i.e. miles away from the top players, which feels about right. Actually it doesn't quite, I should be further away from the top players - possibly the numbers component of the scoring in the system is disproportionately large?

[David O'Donnell]

I reckon David's system as outlined above is as good as any other; my run scores +2 on letters, -62 on numbers and +3 on conundrums for a -57 total, i.e. miles away from the top players, which feels about right. Actually it doesn't quite, I should be further away from the top players - possibly the numbers component of the scoring in the system is disproportionately large?

It could be well be, judging by the scores obtained it seems to have resulted in disproportionately large results.
Maybe 3-2-1 would be better but then I'd also have to lower the conundrum weighting too (and perhaps the niner rating since 7 would now equate to approximately 23.3 pts).

Will try it out.

Also, on a side note Junaid would get -76 on the numbers and Jono gets -77.

[Martin Smith]

I don't like the extra 7 points for missing a nine, it's no different from not spotting an eight in practical terms, and I've always felt that the 18 points is a bit of a gimmick. Also, there's no consideration of whether a risky word wasn't gambled as their opponent had already declared less than their safe backup option - although adding these in would mean fine-combing all 104 games. Maybe extra weighting for crucial conundrums too?

[David O'Donnell]

I don't like the extra 7 points for missing a nine, it's no different from not spotting an eight in practical terms, and I've always felt that the 18 points is a bit of a gimmick. Also, there's no consideration of whether a risky word wasn't gambled as their opponent had already declared less than their safe backup option - although adding these in would mean fine-combing all 104 games. Maybe extra weighting for crucial conundrums too?

The difference in missing a nine is in terms of how it is scored but I did reduce this score to +-5 with +-3 for numbers and conundrums. I also upped the letters to +2.1 and only 0.5 as a bonus for a Darren. This gave the following results for letters, numbers, conundrums and totals. Kirk moved up into fourth but I think that his run was a helluva lot more impressive than this weighting system accords credence.

Craig Beevers -146.3 -48 -14 -208.3
Julian Fell -147.5 -32 -18 -197.5
Chris Davies -134.1 -35 -17 -186.1
Stewart Holden -120.4 -18 -14 -152.4
Kirk Bevins -116.2 -46.5 -9 -171.7
Innis Carson -112.1 -33.5 -17 -162.6
Matthew Shore -109.5 -30 3 -136.5
Conor Travers -107.6 -42 -12 -161.6
David O'Donnell -104.4 -51.5 -5 -160.9
Andrew Hulme -88.8 -43 -16 -147.8
Tom Hargreaves -83.7 -43 -10 -136.7
Jon Corby -80 -40.5 -4 -124.5
Chris Cummins -53 -53 -11 -117

As for risky words I think that overtime a player will incur a disallowed word penalty, I think Andrew Hulme was a good example of a player who wasn't afraid to gamble. Crucial conundrums are somewhat accorded for in that how speedy you were meant you got more points. Though this is not gospel, I tended to do better on conundrums when they were crucial but my brain switched off when they were not. Missing SCORELINE in the final was particularly pathetic but (and I think I can speak for Richard here too) we were both knackered and wanted the whole thing to be over.