The Prediction League (TPL) Competition Rules Review

[bb21]

This is in relation to the main competition referred to in here.

The competition rules review process is now open. Please post your suggestions for improvements below. I cannot guarantee that every suggestion will be implemented however they will all be discussed. Anyone can propose changes. This part is not restricted to those who took part in the first two seasons.

The preliminary review process will finish at midnight on Sunday 15th July.

Any changes proposed will go to a vote by all participants who took part in Season 2. This process will finish at midnight on Sunday 22nd July.

Fixtures selection will take place after that.

Fixtures for Round 1 will be released at least one week before the start of the new season, which is expected to take place on Friday 17th August.

-----

Edit: List of items to be decided by a vote (updated continually)

• Away goals declaration to match submitted match predictions, or left as a free entry;
• Five-division structure (with either SPL or BSP), or four-division structure;
• Attendance precision bonus for percentage margin or absolute value margin;
• Precision bonus value;
• LMS (as a reminder for me, not for public vote );

 

[Ivo]

I agree with the proposal made in the Euro 2012 thread to insist that players' goals predictions should match their fixture predictions, although we will need to be careful in that miscounts can happen so gentle nudges in such situations would help.

Elsewhere, bb21 will remember a suggestion I had a while ago concerning attendances. Rather than giving five-point bonuses to anyone who is 1,000 or less away, it should be done by percentage. A 250,000 round would in theory be easier than a 400,000 to declare. I would recommend a value of 0.5%, although would be happy with any value between 0.25% and 0.5% inclusive.

 

[SS4]

I agree with the proposal made in the Euro 2012 thread to insist that players' goals predictions should match their fixture predictions, although we will need to be careful in that miscounts can happen so gentle nudges in such situations would help.

I'd be happier with a +/- 10% (rounded down) system than an exact match. Say the sum of predicted goals was 23 then you could say anything from 21 to 25. If it were 30 then you could say 27 to 33.
Just to clarify for next season: is the goals scored and attendance based on the matches that crop up in the prediction league itself or for the whole football league?

 

[Ivo]

If it was the whole league there would be some instances of values in excess of 750,000 :?

Speaking of which, I know it was made in jest before, but from a serious perspective could we include the SPL as a competition? If nothing else, if we use it instead of the BSP we won't need to worry about weather problems as often!

I'm not actually advocating dropping the BSP in favour of the SPL. But could we introduce the latter? Or would it realistically require more than 20 matches?

 

[bb21]

I agree with the proposal made in the Euro 2012 thread to insist that players' goals predictions should match their fixture predictions, although we will need to be careful in that miscounts can happen so gentle nudges in such situations would help.

That is a possibility, however in that case a declaration will not be compulsory as the total can be calculated from the predictions themselves. I believe this one can go to a vote, as long as we don't hear too many objections beforehand.

Elsewhere, bb21 will remember a suggestion I had a while ago concerning attendances. Rather than giving five-point bonuses to anyone who is 1,000 or less away, it should be done by percentage. A 250,000 round would in theory be easier than a 400,000 to declare. I would recommend a value of 0.5%, although would be happy with any value between 0.25% and 0.5% inclusive.

I don't have an issue with that, however this would in practice make things more difficult when compiling scoring for me (as I now have to use a calculator rather than relying on my head). Ideally I would like to keep things simple.

An alternative is to declare attendances in 000's, and final attendance is rounded (up, down or to the nearest 1000).

I'd be happier with a +/- 10% (rounded down) system than an exact match. Say the sum of predicted goals was 23 then you could say anything from 21 to 25. If it were 30 then you could say 27 to 33.

Would that make it too easy to get the precision bonus? Five points is a quite large bonus, so it is ideally given only very rarely. Alternatively a smaller bonus could be given.

Just to clarify for next season: is the goals scored and attendance based on the matches that crop up in the prediction league itself or for the whole football league?

All will be the same as for last season unless it is explicitly changed.

 

[Ivo]

I don't have an issue with that, however this would in practice make things more difficult when compiling scoring for me (as I now have to use a calculator rather than relying on my head). Ideally I would like to keep things simple.

An alternative is to declare attendances in 000's, and final attendance is rounded (up, down or to the nearest 1000).

I'm not sure how the bonus would work though. I think that would make it too easy. Alternatively, we could do it to the nearest 500, and then use the following system for the bonus: You must be within 500 to earn the bonus. For every 100,000 spectators the permissible bonus range increases by 500 (e.g. 432,952 would be 430,500 to 435,000).

Or, otherwise, if the attendance is that big a concern... let me do it

 

[bb21]

I'm not sure how the bonus would work though. I think that would make it too easy. Alternatively, we could do it to the nearest 500, and then use the following system for the bonus: You must be within 500 to earn the bonus. For every 100,000 spectators the permissible bonus range increases by 500 (e.g. 432,952 would be 430,500 to 435,000).

Or, otherwise, if the attendance is that big a concern... let me do it

As a proportion or as an absolute value are both fine. I don't like the "stepped" structure as that would really complicate things. Rules need to be simple enough for anyone to understand.

I will have a job for you (not attendances though because I want to keep certain things as they are), as there is something on the side that will be run as a trial from this season onwards.
--- old post above --- --- new post below ---
I don't mind having the SPL onboard, however I don't really want the structure of the game to change so that the SPL will be taking the place of Blue Square Premier if it is taken on.

Also, because we only have 8 games to distribute amongst all other divisions in most cases, I don't really want to spread these too thinly so a six-division structure is out.

Another (less preferrable than the existing arrangement) alternative is to simply get rid of Blue Square Premier altogether and have a four-division structure. I will be willing to compromise on that.

I will leave the decision to the vote.

 

[Ivo]

I will have a job for you (not attendances though because I want to keep certain things as they are), as there is something on the side that will be run as a trial from this season onwards.

Why does this make me feel slightly nervous?

 

[newbie babs]

its way too confusing

 

[Clip]

its way too confusing

Just a bit. normally in other prediction leagues I use its

Result
Correct score
And you have 1 special per week which if right result and correct score will get you a bonus 5 points.



Cant see me entering this if its going to be into this much detail im afraid - I just dont have the time nor inclination.

 

[Ivo]

In relation to it being "confusing", I would have to suggest otherwise. It's really very simple, because all you have to do is declare a series of results and then a couple of other special statistics, and that's it. So there's very little that participants actually need to do - unless of course you're adjudicating. That's a different matter. During the Euro 2012 Group Stage it took me anything up to 40 minutes per day to prepare tables and analysis for the competition - and that's assuming the table is completely error-free! I guess that's what you get for doing it in Notepad, whereas Excel would take about two minutes :roll:

I assume bb21 has a different way of doing things, but either way even this system can be incredibly simple if managed correctly.

 

[bb21]

its way too confusing

I have bullet-pointed a summary of things that will be going to a vote.

The only major change currently being suggested is the criteria under which the attendance precision bonus (currently 5 points for being within 1000) is awarded. All others are minor structural alternations of the format.

One of the proposals is changing it to a dynamic margin (rather than a fixed margin, currently 1000) which is determined depending on the actual total attendance (calculated as 1% of that or 0.5% of that).

I have reservations as mentioned before however I will leave it to a public vote.

Cant see me entering this if its going to be into this much detail im afraid - I just dont have the time nor inclination.

I borrowed bits from a few different versions I have seen.

If you don't know how it works, the easiest way to understand it is to look at some past entries. Reading the rules (most of which relate to procedures and technicalities) by themselves will not be very helpful.

The reason this competition has two tables running concurrently is so that people who miss some entries can still have a chance of winning.

During the Euro 2012 Group Stage it took me anything up to 40 minutes per day to prepare tables and analysis for the competition

Currently with seven active players, I dedicate about 30 minutes per round to scoring the entries and another 15 possibly to sorting out the tables. This includes checking for errors and similar tasks. I only write a short summary of anything notable.

I can't always afford to spend more than an hour on this so that would probably be my limit in a sense. This means that we can accommodate a good few more active players, however that is provided that most of the tasks related to scoring remain reasonably simple.

 

[Clip]

I borrowed bits from a few different versions I have seen.

If you don't know how it works, the easiest way to understand it is to look at some past entries. Reading the rules (most of which relate to procedures and technicalities) by themselves will not be very helpful.
.

Yeah I read the rules and all that but I still feel its over complicating things when you are going on about crowd attendance and all that malarky. Some us only venture that far down the tables when doing a coupon and looking for value bets so would have no idea of average crowd attendances for games down there so would be extra effort to start reading up on it all to get an idea for putting in a prediction.

 

[bb21]

Not everyone's cup of tea granted, and I don't blame ya.

I don't think many of us do have a clue in terms of things like attendance. Sometimes the guesses are wildly out.

 

[Greenback]

I'd like to keep things as simple as possible. It's meant to be fun, not a statistical challenge! At leats with an actual margin, like 1,000, everyone can see at a glance if they have got abonus, making it into a percentage would result in anyone interested having to get their calculator out!

 

[Ivo]

I'd like to keep things as simple as possible. It's meant to be fun, not a statistical challenge! At leats with an actual margin, like 1,000, everyone can see at a glance if they have got abonus, making it into a percentage would result in anyone interested having to get their calculator out!

Then maybe you guys need maths lessons

I guess part of the reason I prefer the percentage method is that I know I can work my results out very easily, even without a calculator. But, if others would have reservations about that method, then keep the current one. Just remember what I have previously said about differing attendance sizes - in theory, players should be closer to a 250,000 total than a 500,000 total.

Alternatively, could we change things to have two levels of bonus points? In addition to the current five for being within 1,000, could we do two for being within 2,500? And then, for away goals, two points for being one off?

 

[Temple Meads]

I'm confused :P

 

[Greenback]

Then maybe you guys need maths lessons

Well, after a hard day at work, I don't relsih the thought of trying to decide if my prediction of 379,452 is within 5% of the actual attendance of 360,871 or not!

I guess part of the reason I prefer the percentage method is that I know I can work my results out very easily, even without a calculator. But, if others would have reservations about that method, then keep the current one.

Excellent!

Just remember what I have previously said about differing attendance sizes - in theory, players should be closer to a 250,000 total than a 500,000 total.

Is there any statistical evidence that shows this to actually be the case?

Alternatively, could we change things to have two levels of bonus points? In addition to the current five for being within 1,000, could we do two for being within 2,500? And then, for away goals, two points for being one off?

I could support this idea. I could also support different amounts to 1,000 and 2,500...

 

[SS4]

I still think 1,000 is too small a margin. For example for a crowd of 250,000 then 1000 is a margin of 0.4%. I could understand if a bonus (say 12 points) were to be given for being this accurate, over the course of a season I can't see it being damaging.
Percentages wouldn't be a bad idea but that seems a lot of work for bb21 and would favour the small attendances although I'd support a tiered bonus system based on absolute divergences from the actual number.

I would prefer to guess the number of cards. Either a sum total like Ivo has done for Euro 2012 or the number of yellow and/or red cards.

in the Euro 2012 thread to insist that players' goals predictions should match their fixture predictions, although we will need to be careful in that miscounts can happen so gentle nudges in such situations would help.

Would away goals follow the same rules?

 

[Ivo]

Well, after a hard day at work, I don't relsih the thought of trying to decide if my prediction of 379,452 is within 5% of the actual attendance of 360,871 or not!

Sorry - it's about 0.015% out

Is there any statistical evidence that shows this to actually be the case?

Not to my knowledge (I haven't checked though); it's a theory based on averages. If all guesses were within 10% of an identified value, then a smaller requirement would be more likely to produce results within a set range.

If we have two values, one of 200 and one of 500 (both of which are unknown), and say that we will give extra bonuses to anyone within 5, if all attempts are within 10% in theory each guess has a 2.5x better chance with the 200 value than the 500 value. 200 would allow 180-220, and especially 195-205 (roughly 1-in-4 if all were within 10%); 500 would allow 450-550, and especially 495-505 (roughly 1-in-10).

I would prefer to guess the number of cards. Either a sum total like Ivo has done for Euro 2012 or the number of yellow and/or red cards.

Cards can be difficult to do if the source cannot be confirmed to be accurate (and using the BBC site has proven very difficult for Euro 2012). However, I would suggest an absolute number might work; a typical 20-game round should produce around 70-80 but would see some fluctuations which would be ideal for this game, especially considering the teams (outside of the Championship) involved. However, unlike away goals and attendances there is a lot more guess work involved; at least attendances can be estimated from stadium size, fan base and match prestige. Overall therefore, I think cards as a bonus are better suited to smaller competitions, such as the Euros.

Would away goals follow the same rules?

Probably.

 

[Clip]

Then maybe you guys need maths lessons

Or after a hard day at work trying to rectify other people mistakes on either changeovers or refunds I really dont want to be doing any more maths thanks very much.

So on that final note ill count myself out of next season.

 

[Ivo]

I hope I haven't put you off :? Sorry if I have!

You'd be more than welcome to join at a later stage if you wanted to. And (to everyone) remember that only the adjudicator has to do the hard work!

 

[Clip]

Nah dont worry about it. I like the simplistic things in life

I can understand why some of you like the whole detail about stuff like this but its not really my thing overall. I have a hard enough time trying to get value out of the weekend coupon

 

[Greenback]

Can't we just leave things as they were?

 

[bb21]

Of course, we can. There will definitely be an option to leave things as they are.

 

[newbie babs]

As the person who was the winner of The Prediction League last time. I think you are making it complicated and maybe I am simple but I like simplicity and its getting to be a complicated for those that are not into stats and maths and I am not into that.

I would therefore rather not take part. I don't understand how it works and I just want to predict scores so have fun in your challenge.

 

[valenta]

Would agree with Newbie Babs, I think the new maths element could make the game a little too complicated. Though often relying on the bonus points myself, I feel that the actual "prediction points" (from the accuracy of the score line) are the most important. Yes, the bonus points are a nice extra - but I feel we should keep these simple, partly due to how much my maths skills would be comprimised!

 

[bb21]

Don't feel pressurised into accepting the proposals. As I said, feel free to vote for things to stay as they are. This is just an opportunity for people to put any ideas they might have forward.

 

[Ivo]

I first had the percentages idea a few months ago and, whilst I would prefer such a system, if others would like to keep things as they are I'm happy for that to happen; after all, it's a pretty minor part of the game. Maybe maintaining the exact system used in Season 2 would be best then?