At week 8, it's probably time for a look at the judges' scoring to date. What I've done is, for each judge in turn, compare their scores for each couple against the average scores from the other judges:

As expected, Craig's figures are lower than for the others, and tailing off as you go down the list. Conversely, the other judges' figures are nearly all greater than 100%, and rising gently as you go down the list (to compensate for the tailing off of Craig's).

Next to look for are figures implying bias. Craig seems to be down on Simon. Darcey is a bit down on Frankie, and up on Simon and Steve. Len slightly favours Frankie, and the rising characteristic shows that he is generous with his scores for the weaker celebs. Bruno is maybe slightly down on Jake, but all his scoring is rather on the generous side.

Darcey seems to favour the male celebs, Len and Bruno the female celebs.

(Thanks to ultimatestrictly.com for the base data - though they made it a bit harder this year by including Donny's scores, so I had a bit more work to do.)

Feel free to comment!

Craig tends to use the full range of marks available to him and the others don't*...with the result that the weaker the dancer the more their total score is artificially inflated. I understand why they do this and, to be fair, it's only the order on the LB that really matters not the actual score.

*I'd be very interested to see the comparison of how many times each judge has marked under 5.

That would take a bit of work, using ultimatestrictly.com, but they've got info regarding the total number of awarded 1's: it's 9 (8 from Craig, 1 from Arlene):

http://www.ultimatestrictly.com/the-lowly-one/

Wow you put a lot of work in, thank you.

Craig marks with a greater range than the others, that much we know already. What's interesting is the variation from simple proportionality. As you say. Bruno favours Pixie and Caroline, not so Steve. Len generally average marks, but Mark and Simon miss out. Darcy marks Frankie and Suntetra lower than the others.

One analysis I never got round to is "how often is Craig's mark the only separation on the leader board?" What I mean is there's an established factor of ties bringing the middle of the leader board into dance off by bunching scores. If it wasn't for Craig using more of the paddles, most weeks you'd have many more on all 7s, all 8s and so on.

Thank you for doing this OP
Im convinced that Craig has something agaisnt Simon ; Crazy how he tends to rate Mark higher than him when IMO Simon is the best out of the 2.

You're welcome Terry - actually though it wasn't too bad (about half an hour's work?).


Yes, that's a good point - we already have too many ties, it would be worse if it wasn't for Craig. (I wish the PTB had chosen a less ridiculous "improved" points allocation system, but that's a different matter.)

Great work - it suggests some interesting trends.

What about Donny? Was he really as bad as he is remembered or was the Donny 10 unrepresentative of his overall scoring profile?

I guess it all depends on whether they are actually marking based entirely on their own opinions.

Then upon whether they are marking based on dancing or entertainment.

Then upon whether there is any logic involved at all.

Thanks for doing this OP.


I wonder if Craig marks Simon lower because he thinks he is really very, very good and wants him to improve even more, rather than out of favouritism for Mark.
I like to think that anyway.

Out of all the judges, I think that Craig is the most reliable overall, but even he can get swayed (whether by the moment or the need to balance out the voting figures, who knows?).

I don't think those figures do imply bias. A difference of opinion doesn't have to mean bias. Indeed, it could be the other judges who are biased and the judge who's score is furthest from the average was the only unbiased one.

But it's also not clear how you're using the figures to find "bias". What, for example, means that Craig is "down on Simon" but not on Steve or Judy? Craig's score for Simon was 81% of the average, but it was also 81% for Steve, and 57% for Judy.

Scoring wise Craig lost all credibility when 'funnity fun fun fun' was a contestant.

The others never had any.

Still, as others have said, as long as the order on the leaderboard is roughly correct, it doesn't actually matter much.

I agree, but only up to a point.

It doesn't matter all that much whether their scores for the weaker dancers are "artificially inflated" because the leader-board positions are based only on relative values. I wonder how many of Craigs 5 and lower marks made a difference to the leader-board.

Craig doesn't use the full range of marks; he just (usually) uses a wider range than the others. But that's largely because he gives lower marks to people who were going to be low on the leader-board anyway. (His range is usually more restricted than the others' at the top end.)

In any case, the range of marks Craig uses is often greater than the number of different marks he uses.

If we consider the number of different marks, Craig's not so different from the others. In the Blackpool marks, for example, Craig's used 5 different marks, but the three other judges each used 4.

If we set aside Craig's 4 for Judy (which made no more difference to the leader-board than if he'd re-used one of his other low marks), the difference between Len and the others would be that Craig used 6, 7, 8, 9 rather than 7, 8, 9, 10 like Len and Bruno, or 6, 8, 9, 10 like Darcey.

Next ... ties:

I was confused by Craig on ITT yesterday when he said had he known Judy would be going out he would have given her a 5 but he was convinced she would be staying in so gave her the 4. So which did he think she deserved? I can't see how it made that much difference but that sounded like tactical voting.

"Bias" technically just means "inclination". You can investigate or suggest why there is an inclination.

I did something similar a couple of years ago. Craig's stats for Lisa were interesting. Nothing similar this year.

Also similar stats for Louis IIRC were remarkably similar to what has been produced here for Simon.

Beginning to look like a scripted story line. Awkward reserved blokey bloke comes out of his shell to become a Daahling Daancer and lifts the cup at the end of the season.

He gave Judy the highest mark he could to ensure that Judy came last on the judges' board; yet not generate a voter backlash support for Judy. Otherwise he would have given her a 3.

All scripted by the producers anyway.

That Craig uses a greater range of marks does not mean he contributes less to ties.

For example, in Blackpool, when there were three ties, one of the reasons Simon and Pixie were tied was that Craig gave them different marks (8, 9) that balanced the different marks Darcey gave them (10, 9).

The same thing happened with Frankie and Caroline's tie, except that it was Len who gave the (10, 9).

So what about the 3rd tie, between Jake and Mark? Craig was in there too, because all of the judges gave both 9.

There used to be a similar view that Craig's marks made more difference to leader-board positions than the marks from the other judges, but a closer look showed that wasn't necessarily so.

Indeed, the best way to make a difference to the leader-board is to disagree with the other judges: the sort of thing that's being labelled "bias" in this thread.

What is the reasoning that makes 4 the highest mark he could give to ensure that Judy came last? She'd have come last if his mark had been anything less than 10.


You can investigate or suggest why there is a difference in marks. Calling it an "inclination" is already suggesting an explanation.

Re, "technically just means", that's ignoring everything that makes one choose to use the word "bias" rather than "inclination".


What was interesting about Craig's stats for Lisa?


What is your theory re scripting? Are the judges told what scores to give? Just told who to favour, with the scores they use left up to them?

Yeah, I forgot that.
A bad time.....

Can't speak for Henrywilliams, but for me, that sounds like Craig's personal script (if it is indeed a script), rather than the producers'.

It was mooted in 2012 that Craig needed her kept in for as long as possible as he wanted her to host his tour which wouldn't have been feasible if she'd been eliminated by week 2.

He must have a script.

Glad we've cleared that up then😳😳😳

One thing I noticed is that the only celeb where anyone other than Craig is less than 100% of the average is Darcey's marking of Frankie. Does she not like her (or does she not like Kevin...).

She definitely seems to have a downer on her.

Other than that, pretty much what you think is happening, but nice to see it layed out before us.

Good work VintageWhine.

I think Mark fits in more with Craig's dance aesthetic- he loves an OTT, melodramatic, jazz-handy performer. Simon is a tad too subtle for him by comparison.