View forum thread

Maths is for dummies
Yesterday, my average squad rating was 69. I sold two players, both rated 57.

Today, my average squad rating is 68.

Come on FS.

Isn't it bleedin' obvious?

Yes. Either I need to redo my maths degree, or IAG needs to take one.

If you remove 2 data points from a set of numbers that are below the average of that set, the average should go up. Not down.

Those were two of my lowest rated players. Sigh... I think it's time to give up on trying to work out how APR works...

I would be confused by that too, but the game has a habit of updating some functions before others & that may well be the case here.

For instance, a loss in a match generates an instant drop in club &/or manager rating, but getting a better member of staff isn't taken into account during that update, so you have to wait until the following update to see the benefit of your new member of staff.

That may not fully explain your confusion, but is the most logical reason for the skewed figure in my opinion.

No, BlueBoar, I specifically waited until the update after I sold the players to check my APR. They sold in the 8pm Wednesday update. I lost a point of APR in Thursday's 8pm update. None of my other players have lost any rating and several are improving. This has nothing to do with the ordering of the update. This has exclusively to do with the indecipherable way that IAG works out "averages".

See, I thought that the APR calculation discounted some of your lowest rated players - I worked on removing two for a while but it could also be possible that in larger squads that number is increased to three or even four. Maybe if you moved over a breakpoint a particularly low rated player is now included in the calculation?

Here's a (not necessarily the) calculation on your squad...

Average of the top 22 players in your squad now = 67.8

That includes 50-rated Mahmoudi

Add in two 57-rated players and the average of the top 22 = 68.2

Now let's say every X-rated player is actually X.49...the averages become 68.29 vs 68.69

Round them and Bob's your uncle

So what you're saying is that there's a possibility that the game only calculates the best X players in your squad, regardless of squad size? (In this case with X = 22?) That seems quite unlikely to me and reasonably easy to test by simply promoting an additional youth. I did so towards the end of last season (Kelly) and once again my APR dropped from 69 to 68- it then returned back to 69 a few days later. That would seem to disprove your theory as my best 22 players didn't change during that time.

Open to other suggestions, including that IAG simply doesn't know how to add up or divide...

That's what I've heard before or rather that it cut off the bottom few. Like I said, maybe you'd moved over a breakpoint in the number of players it discounted. That approach makes sense to me as it allows you to bring in some youth players without massively affecting your APR.

Selling 2 players moved over the break point? OK, let's argue that point for a moment. There's no way that going from a squad of 28 to a squad of 26 changes X by zero, right? Otherwise I could now promote 2 new youths with no downside at all. It would make more sense for X to have changed by one, which is why you're referencing the break point, right?

That means that instead of counting 2 players rated 57 in the calculation, it's counting 1 player rated 50.

But that makes no sense either because one player rated 50 is dropping 19 points from my average previous APR of 69 while two players rated 57 were dropping 24 points from my average APR of 69. The total sum of the players being counted would actually have increased, meaning that the average should go up.

Only if the game is also counting Holme, rated 44, as well (that is, if X is indeed changing by zero) does it make any sense at all, and if that's the case, I'm now free to promote my next 2 youths with no adverse effect on my APR. But does that tally with other peoples' experiences? I highly doubt it.

I don't think there's a mathematical way out of this one that also makes sense in the context of the game and previous observations. PS seems to think it's bleedin' obvious but he's not sharing if he was indeed at all serious..

Alright then, let's say that having 26 players in your squad gives you three 'free' players and having 28 gives you four.

Average of your current top 23 players = 66.78

Average of the top 24 players if you add in two 57-rated players = 66.92

So you've added two players that are below the average but the average goes up.

Of course this doesn't track exactly with your observed APRs but there might be more going on. Maybe IAG gives your top 11/16 players a higher weighting?

Or the exclusion could be done by percentage of your squad rather than raw number. Or some kind of "number/% below total average" calculation. Lots of options.

Ultimately, I think it's fair to say that removing players below your APR and adding ones above it will in the long run improve it even if there are a few edge cases that might look a bit funky in the day to day changes.

Well FS, did you get your rating back?

Nope. Still on 68.

Hum, very strange. Commiserations

Doesn't really matter. I'm going to have to get to about 75 to get the next staff level anyway, so a point here or there isn't really important right now.

It does help a little, though. The key seems to be to have a larger squad to increase your APR, even if the players are slightly lower quality.

Hah!

I know this will probably roll back next update, but this makes me laugh.

I sold an old GK today who was rated 77, expecting to go back down to 68 ASR once again.

Instead my ASR has gone UP to 70.

The solution, therefore, is obviously to sell all my best players and keep my worst ones

You should continue the experiment therefore in order to test your theory on more than one example.

I notice PS isn't here saying that it was bleedin' obvious that this would happen, and chyper hasn't offered his expert mathematical advice on this one either

Now you wouldn't be getting cynical, would you, FS?

Sheesh, some of us had Olympics closing ceremonies to watch you know

Alright, I'll admit that one is a bit odder. So either Ron manually picks APRs every update specifically to annoy FS, or the formula for calculating APR is more complicated than a straight average.

Did the rating go back down again this update?