Forgive me for not including this piece in the first DPL Article, but we need a little time to try to reach agreement over the scoring system. From the discussion last season it does seem that we generally believe that 5 points for guessing the size of the crowd is disproportionate to what we might loosely describe as the test of predictive skill which is the essence of the competition.
We could just eliminate the crowd from the list of predictions or we could reduce the points allocation or even narrow the scoring band. An alternative might be to replace it with the half time score. We might also perhaps revisit the points awarded for guessing the various outcomes, the score, the result and the scorer. Retaining the 10-point bonus for a 100% accurate forecast is probably the one element that would be universally agreed.
I find the science of statistics somewhat baffling so my simple views are not up with the professionals in the subject. However, it does seem to me that the actual score is more difficult to predict than ether the scorer or and the result which obviously has only 3 options and is therefore easiest of all.
If we did want to retain the crowd element perhaps 3 points and a 100-person margin might be fair. We could raise the actual score prediction to 5, the scorer to 4 and keep the result at 3. Fixing points for a half time score seems the most challenging statistical task, so might best be avoided.
DPL Points system for 2019/20
posted on 28/7/19
Keep it simple. Result, score and scorer. With perhaps only a 5 point bonus for a full house. Assume predicting no scorer gets the points too?
posted on 29/7/19
I agree with most of the above:
Drop the crowd prediction (even though I got many points from this!);
3 points for result, score, and scorer;
Reduce the bonus to 5 points;
However, if one predicts Rovers will not score, one SHOULD NOT get scorer points - that used to be called "double-dipping"!
I've been travelling over several weeks and it is why you have had no comments from me about what has been going on at the Keepmoat. While I agree to a certain extent with what Hound (I think) said about people wanting to move to a higher level, personally I was disgusted by what GM did - losing many of our better players and then quitting. I'm sure glad Hull did not win on Saturday. That performance by our lads was the first encouragement I've felt in a few weeks.
posted on 29/7/19
Micky and Nookie: So far then, the crowd is removed.
The question of scorer points for an nil score leads me to admit that last season I did award 3 scorer points when 0 was successfully predicted.
I knew nothing of “double dipping” until today and have attempted to acquaint myself with the term which is used in Statistics, apparently when you over-ascribe value to a number.
In this context one may predict 0 goals, but you are not actually naming a scorer, so on that basis do not warrant the 3 points. Admittedly you cannot do so, but if you predicted 0 for each game, you could probably collect more points than consistently naming John Marquis as the scorer of the first goal.
However, if you do get the result right, even if you have not named a scorer (because there was none) I believe that this should not also disqualify you from getting the bonus.
Would that be fair?
posted on 29/7/19
Getting double points if 26 scores would be good.
That would make a big difference to the points tally
posted on 29/7/19
No it should be 26 points if 26 scores!
posted on 30/7/19
Donaldo - I agree that the bonus should go to someone who predicts zero for the Rovers and gets the score correct.
posted on 31/7/19
How about removing JM from scorer predictions.
posted on 31/7/19
Predicting no goals for Rovers now looks like a good bet!