Announcements
*** June Competition Winners: Well done to Le Male (NAPS), Coco69 (KO Cup), Saddlesore (Royal Ascot) & Like2Fish (Poker) ***
*** New Last Man Standing Competition - Win up to £1500 Annually - See Competitions Forum ***
Sir Puntalot

2018 ELO Ratings Development

Recommended Posts

So I thought I'd share with you what we have planned for 2018 for the ELO Ratings. :ok 

Firstly, there's a few leagues we will drop because the data is not always there and that's no good. We will settle on a good number of leagues to be announced soon, obviously all the UK and major European leagues are in that without question.

The first thing we'll be doing is upgrading the design to be far more mobile friendly and with odds included for Prediction Ratings, this is not as easy as it sounds as mobile is very small, but I think between me and our designer, we've done a good job.

Performance Ratings will also get the same update, and the idea is to include the Prediction Rating alongside the performance ratings, so at a glance, you can see whether that result rating matched the initial prediction.

The more juicy part I want to get to, is making the ELO Ratings produce football tips for all the leagues we cover. This is a big task and one I have to chat to our developer about, but I'm sure we can get it done towards the end of this year.

This would mean that a certain range of ratings would mean X. For example, it would have to work on the last 2 average (makes sense to do that as it's a fresher rating) and that a range of perhaps -9.99 to +9.99 would predict a draw. Now, this is yet to be finalised by myself, but that's the idea. Then -10 would mean away win, +10 home win etc...

We can then advise accumulators on the strongest selections and even teams we think you should be laying, as well as value longshots, BTTS, Over 2.5, I could go on forever but I'm sure you see the potential. It would effectively do similar to what we have for horse racing tips now, however compiling results for all those matches might not be possible unless we can automate it. We'll see how we get along. :ok 

There's a lot we can do, but it requires integration of our odds development, the tips system and the ratings, so it's not a simple task but I do believe we'll nail it somehow. ;) 

 

Share this post


Link to post
Share on other sites
2 hours ago, Kentucky110 said:

Hi All

Would be nice if there was an option to view new style and old style ratings

Even better would be to have the ability to download both prediction and performance ratings

 

@Kentucky110 - I assume you mean to have the option to switch to how they are currently? Not sure that will be the case, but they'll be better especially on performance ratings side.

Download ability - I'll put that on the list. :ok 

Share this post


Link to post
Share on other sites

Good afternoon, first of all congratulations for the beautiful system, I'm really fascinated. Still giving a studied (recently) and verified that the forecasts are much more successful when there is a difference greater than 10 for both the 4 avg and 2 avg situations, would this be a safe parameter for betting? Or would a greater difference in both be more reliable, for example, above 20 for 4 avg and 2 avg?
Already grateful for the clarification,
Regards,
Artur.

Share this post


Link to post
Share on other sites
3 hours ago, artursfas said:

Good afternoon, first of all congratulations for the beautiful system, I'm really fascinated. Still giving a studied (recently) and verified that the forecasts are much more successful when there is a difference greater than 10 for both the 4 avg and 2 avg situations, would this be a safe parameter for betting? Or would a greater difference in both be more reliable, for example, above 20 for 4 avg and 2 avg?
Already grateful for the clarification,
Regards,
Artur.

:welcome to PL @artursfas  :ok 

It's a difficult question to answer for various reasons. Obviously the bigger the difference, the better chance that team should have over the opposition. However, with all ratings they can only ever be a guide, but they are at their most powerful when only a few matches have been played and towards the end of the season, because bookies decrease the odds of motivated teams over teams that appear to have nothing to play for. ;) 

Share this post


Link to post
Share on other sites

I'd like to say thank you for the ratings Sir Puntalot.

I have been looking into some systems and think the prediction 4Av , 2Av and differences ( and the difference of the differences) may be a useful filter.

Still very much a beginner and collecting lots of data, but I'll let you know if it comes to anything.

Downloading as an excel sheet would be excellent.

I have no idea how you do them, but as a thought experiment, I'd say % possession, shots on target , shots off target, corners etc

Must be very time consuming, so it's very much appreciated.

 

Cheers

Share this post


Link to post
Share on other sites
6 minutes ago, Rabsterman said:

I'd like to say thank you for the ratings Sir Puntalot.

I have been looking into some systems and think the prediction 4Av , 2Av and differences ( and the difference of the differences) may be a useful filter.

Still very much a beginner and collecting lots of data, but I'll let you know if it comes to anything.

Downloading as an excel sheet would be excellent.

I have no idea how you do them, but as a thought experiment, I'd say % possession, shots on target , shots off target, corners etc

Must be very time consuming, so it's very much appreciated.

 

Cheers

@Rabsterman  Glad you like them. :ok 

Plans are in motion to make them even better, but development takes a while sometimes. :eyes 

I've noted everything here, but it will be towards the latter end of the year when we should make real progress. We have to move to a new setup that will give us much more freedom to make advances across the site as a whole. :ok 

Share this post


Link to post
Share on other sites

Hi sir puntalot 

I have to commend you for your work on Elo rating and your proposed plan of using it as prediction tips. 

I will like you to incoperate (scoring strenght) last 5 games average goals score for home and average goals scored for away team and also looking at current position. With this you can get a better prediction because you considered the basics of football. 

can wait for next season. 

Regards 

Share this post


Link to post
Share on other sites
57 minutes ago, AC23 said:

Am I wrong in the following assumption?

A team can have a strong 2/4 weekly average due to paying/beating mediocre teams.

I've noticed that occassionaly the ratings can get it very wrong and I think this is why?

Like today with the Coventry Exeter game.

I know it's a rough guide and not an exact science, but maybe score a teams performance against their opponents worth? Maybe according to their relative position in the league?

Ie a win against Stoke at the end of last season is worth only 60% as opposed to 100% against Man City.

 

You're spot on. :ok 

I think any ratings can be very wrong occasionally and that's not a defensive answer, it's just a fact. I've often thought of trying to implement what you said above, but it complicates things more than it solves them.

Example A:

Let's say West Brom have a 4 home game average of 35.00 but a 2 home game average of 7.50

Results show their last 2 games were against Man City and Liverpool, but the last 4 were Man City, Liverpool, Stoke and Southampton - you would be right in saying, we should "weight" that because it's obvious they're going to be worse against Man City and Liverpool right?

But is it obvious?

Example B:

Burnley have a 4 home game average of 42.00 and a 2 home game average of 35.00, yet they've play exactly the same teams as West Brom did.

You see where I'm going with it? Ratings should be simple, but logical and that's how I created them. The majority of the time they are alarmingly accurate, but can only ever be a guide.

 

 

Share this post


Link to post
Share on other sites

@AC23 - Maybe I didn't explain that very well. History wise, I think it's fair to say on average West Brom and Burnley would be about the same level over the years.

In the theoretical examples above, Burnley managed to put in a good performance despite the teams being top 4 status, but West Brom didn't so weighting the ratings because of league position/club strength (this is very debatable too) them would just complicate the ratings significantly and probably make them worse.

I think it's a case of you have to dig into the ratings, and check who they played to perform those figures but they serve an "at a glance" guide. ;) 

Share this post


Link to post
Share on other sites
Just now, AC23 said:

If yes - I'd say the way to do it is that Burnleys trying should be relatively unhanged/affected in the games against the top 6, but bottom 6, they get marked down.

Sounds logical, but what would you have done when Chelsea were bottom 6 in that disastrous spell Mourinho had there? You see how complicated it gets?

Share this post


Link to post
Share on other sites
8 minutes ago, AC23 said:

But if a team skyrockets because they hammered the two bottom of the table teams recently, a check in place for that ensures you avoid...you know what I'm say saying.

They still have to perform though. ;) 

I remember a few years back an example of Celtic putting in a couple of performances way below par, but getting wins. Their opponents had been playing well, yet had 1 loss and 1 draw. I layed Celtic at 1/6 and everyone thought I was on a cocktail of acid/mushrooms, but Celtic drew that match 1-1 (late equaliser) and performance ratings after backed up they were again well below par.

@AC23 - I do take on your points though and will have a good think about it. :ok 

Share this post


Link to post
Share on other sites
3 minutes ago, Mindfulness said:

Can we not mess with the ELO ratings please, they're fine as they are in my view. If you want to add deeper layers of context then just do your own research as individual punters. No need to convolute a ratings system which has proven its worth on this forum. :ok

I've had a think about it and they won't change, so no worries there. We will add corner ratings later in the year, amongst the 8 million other things being built. ;) 

Share this post


Link to post
Share on other sites

Any chance you can click on the relevant league and see it for yourself? ;) 

Not right now, the old site is being locked down at the weekend as we move to Saturn, a more glorious atmosphere with lovely rings going round it. So no more updates for a while, until we get to the Mothership.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.