Scouting improvements

Subscribe to Scouting improvements 46 post(s), 15 voice(s)

 

August 01, 2013 22:58

573 posts(s)

Donator

 

It wouldn’t hurt though if the scouting reports are more accurate than random numbers, would it?

 

August 06, 2013 12:47

3,888 posts(s)

Administator

 

From the comments so far there is some preference for more accurate scout reports, even though a few still think that it’s good to have them as they are. I’ve just posted in another topic that maybe it would be preferable to have more accuracy and less scout report requests available. Do you guys agree?

 

August 06, 2013 14:40

573 posts(s)

Donator

 

I think the frequency of scout reports is fine.

Gabriel, do you agree that 20 points difference in two scout reports on a single ability can be called completely inaccurate? Especially since all abilities are in a 30 points range (70 to 100).

Just to check: you did realize when programming that ‘maximum 10 points off the real value’ can give 20 points difference between two reports, right?

 

August 07, 2013 11:38

61 posts(s)

 

I think more accuracy with the same frequency of scouts will make it TOO EASY to find good players…
Actually I think that with more accuracy the number of scouts should be MUCH LOWER.

If we lose the random factor + manager judgement on improving players, I think we’ll lose a little bit of fun.

 

August 07, 2013 13:19

573 posts(s)

Donator

 

Why would it be too easy? You can do 1 report per turn at most, and losing the random factor is something you made up, no one suggested that.

But it’s just stupid to have a feature if it doesn’t work, and if scouting reports are not accurate at all, I’d say that is the case.

 

August 07, 2013 15:06

37 posts(s)

 

Again for what it’s worth, I’m happy the way they are. Adds that “random factor” to the game ad Daniel mentioned above. The scouts failed me with some but also helped me with some. I’m OK that they get it wrong sometimes.

 

August 07, 2013 15:09

37 posts(s)

 

To add, if you don’t think the feature works (I think it does just fine) – don’t use it. You will just play the game prior to the feature and use your own judgement like we all did.

 

August 08, 2013 08:48

573 posts(s)

Donator

 

The formulas should be more in line with common sense and reality
Currently:
higher scout level → more accuracy → Good (better scouts give better results)
Younger player → less accuracy → good (the closer to final level, the easier to predict)
Random factor → in general good, but not well programmed.

One example of what should not have been the case.

22 yr old, renewed scout report. Control is 87, old report said 87 potential level, new report says 100. That is 13 points difference on a scale of about 20, so about 35% accuracy. Too much for a 22yr old (with high scouting levels).

(now my bath is almost full, to be continued!)

 

August 08, 2013 10:14

573 posts(s)

Donator

 

I would like to make ‘accuracy’ concrete. What is acceptable for the random factor?

So let’s say you have a new youth player, shooting 65, and that this means his potential will be between 75 and 95.
As a human, you will probably add 20 points to estimate his potential. So you guess 75, but you know you can be about 10 points off.
This means that 10 points off is a baseline.

First question: at what level should an scout be as accurate as the baseline? I’d say about level 5
Second question: how much more accurate should a level 10 coach be? At the youngest age, a little bit, about 7 points off max.
Third question: how should accuracy improve over age? I’d say 1 point in two years, or 2 points in 3 years. So a level 10 coach on a 22 yr old is max 3 to 4 points off.
Fourth question: should the points off be in both directions for the same player? So let’s say the first scout report on the new player says 75 shooting and the real value is 85, can the second scout report one year later be 95? In my opinion, not. 20 points difference is too much. Maybe 15 (for a young player) is still acceptable? Not for a 22 yr old though, or on scout level 10.
Fifth question: should the real potential value change? I don’t know if it does, but I guess it should. If the new youth player by chance improves 10 points on shooting in his first year, it means that a human scout will have a higher estimate than 85, so the computer scouts should have higher estimates on average too.
Sixth question: should the estimate on all the skills have any connection, or all be random? In other words, if an estimate on potential value of one skill can be max 10 points off, can the average of all the skills also be max 10 points off (so that a player who is 88 average can have a potential value of 78 average), or should that be less? Not sure about the answer, since the chance on that is very low anyway.

I would like to know how the developers and other players answer these questions (and if I forgot something).

 

August 08, 2013 22:20

37 posts(s)

 

I think this is the first time I agree with Dimitri.

I haven’t encountered huge discrepancies with older players (ex. 22), but at scouting level 10 for “Control” 87 jump to 100, I’d agree, a tad ridiculous at the age 22. Unless the first reading took place when the player was 17. I’d be curious to see the difference at age 23.

 

August 09, 2013 14:21

295 posts(s)

 

I would like to see much more accurate reports at the cost of the amounts of times we can scout per season.
IMO quality is better than quantity…

does anyone agree with this? perhaps limiting scouting to 10 times per season or something, but get much more accurate results would be better than being able to scout every turn but get much less accurate results…

 

August 10, 2013 02:09

828 posts(s)

 

how about this …. fuck scouting improve youth development give us a full youth team we can get rid of those we want, now remember theres wage limits n shit, so not every player you want to sign will be able to be signed due to some managers dumb wage offers ( me excluded im on lv1 wage by choice)

the ability to personaly pick areas to improve youth would be better, and yes coaches focus on things to train this game has fuck all coaching imo

read bugs to understand why im pissed off, and tbh i couldnt give a fuck whos offended

 

August 13, 2013 06:03

3,888 posts(s)

Administator

 

Comments on some of Dimitri’s questions:

Fourth question – interesting opinion, do you have that opinion for two different skill results in the same report or only same skill results on different reports? I mean, if a scout report is underestimating the shooting could it overestimate the tackle for instance? Maybe you’re suggesting that the previous report from the same team should be taken in consideration when generating a new report.

Firth question – real potential values change, up to a certain age (the exact age varies a little bit)

Sixth question – they are currently estimated independently, maybe connecting them would be more reasonable

 

August 13, 2013 08:08

573 posts(s)

Donator

 

4: same skill results in different reports. So yes, the previous report should be considered, humans do that too.

 

August 13, 2013 22:43

61 posts(s)

 

I think this is a very good idea!
The next team`s report on same player should make the scout more accurate ( considering the Scouting Level of each report already made on him ).

Maybe this could be a good solution for the report`s accuracy issue!

 

August 14, 2013 03:19

3,888 posts(s)

Administator

 

Yeah…if we limit how off the results are and use any existing reports as the basis we should improve the accuracy without much change and without making it too easy to scout players. I like that.