One of my main hobbies is running "prediction" type games on football that raise money for charity, at my local pub. We (it's me and a couple of mates that organise them) are usually very successful - the one we ran on Euro 2008 raised £175, and as I said on another thread I was embarassed to actually win the thing, this time, myself.

The trick, we've discovered through trial and error, is to follow three golden rules:

1. Make the game appear to be about your superior football knowledge, so you attract lots of entries from every bloke who can't help but be put to the test.

2. Introduce enough random elements into it so that it's pretty much impossible for anyone, despite their football knowledge, to predict things (otherwise you end up with lots of identical "winning" entries"). THe beauty of making a fair bit of it random, too, is that other people spot it's fairly random, so will "have a go" as well (the aim, after all, is to maximise the number of entries and hence money for charity).

3. Make it simple enough to administer that you can keep track of it. I'm not attempting to rival Carling Opta Index here in my study just to run a sodding game.

Now John, one of the guys I run these games with, has had an idea for next season that I think's brilliant. We'll group the 20 teams in the Premiership into 4 pots of 5 (based on last year's finishing positions). Everyone can pick 2 teams from each "pot", so everyone has 8 Premiership teams "playing" for them next season. But obviously no-one can have more than 2 of the big 4, or less than 2 of the expected bottom 5, etc.

Then, the deciding factor will be goals your 8 selected teams score, in the League Cup and FA Cup. Not in the Premiership; that will be irrelevant. Just the Cup competitions.

Now why I see this as brilliant (according to my criteria above) is that instinctively you'd go for the "top" teams, wouldn't you, but then, a) the top teams don't always take the Cups seriously, especially the League Cup; b) no-one has any way of predicting who will draw whom in rounds of the Cup, and c) it's virtually guaranteed to go right to the end of the whole season, well at least to the semi-final stage of the FA Cup, before anyone's entry would be definitely "out". Last year, for example, you'd probably have done well choosing Spurs and Portsmouth, but who would have predicted that?

To make things even MORE random, and really introduce that element of chance, we're thinking of including goals scored in any penalty shoot-outs. For all anyone knows, next year's FA Cup Final could end up a Liverpool 7-4 win over West Ham (if we were to include penalty shootout goals) like in 2006.

What do you reckon, OTF? Sound like a fun game? Would you enter that, £2 to go in, £1 to charity, £1 into the prze pot?

It'c not clear to me what we're actually trying to do - pick the 8 teams that will score most goals?

Presumably one gets one points for wins, league placings, progress in the cups, etc., but I agree that it may be over-complicated to serve the objective (though beer always helps).

The ursus prediction league team found that awarding tripe points for successful score predictions introduced just the proper amount of randomness to keep the competition open to the very end.

Yep, all purely about goals scored in the 2 Cup competitions. Only your initial entry would be "seeded" to an extent from the start, so it'll be 8 teams, but no more than 2 from "the top 5", 2 from the next 5, etc. A bit like filling in a lottery ticket.

So if Bolton draw, say, Exeter, and beat them 5-0, you'll get 5 if you've got Bolton, and if Chelsea draw Arsenal and lose 1-0, you'll get nowt for Chelsea.

Too random for you to get my money.

Right, I think I follow.

Difficult to know how well it's going to work, given that there are only a handful of teams left in cup competitions towards the end, it might be all over bar the shouting well before the end of the season.

Any way of giving it a trial run? (I suppose there isn't really, is there?)