Subject: »[info]DevDiary 80: “Price is what you pay, value is what

people DON'T expect their results based on the budget
people EXPECT PSG to win because of how much money they SPENT on the players that they HAVE

That's true. Maybe this is why PSG will never sell out their players to build team only around 17-19yo youngs.

That proves that problem is not in algorithm itself (for sure it can be adjusted, but still will never fit perfect to any league), but problem is in teams who are selling out their players.
You just can't do it so easy without consequences in new system.

New fans mood bonus/penalty rewards those who are good tacticians, who makes good transfers, who balances better between holding good players and develope young ones.

What are you talking about?

The strongest team in sokker, playing for 82-84, with very wide roster, who won last 2 Ekstraklasa trophies and 2 Polish Cup trophies as well as last Champions Cup, is considered as a team expected to end up a 9th (!) position in league

Championship (1 - 1) KS Wisła Puławy
Vice-championship (2 - 3) WislaKrakow
Vice-championship (2 - 3) MOSP Białystok
Top half of the table (4 - 4) KS Polonia Skierniewice
Top half of the table (5 - 9) ASRIM Iskrzynia
Top half of the table (5 - 9) WHITE TIGERS CF
Top half of the table (5 - 9) FC Chełmek
Top half of the table (5 - 9) Japanese Futbol
Top half of the table (5 - 9) Domik Tuchola
Avoiding direct relegation (10 - 12) Patataj
Avoiding direct relegation (10 - 12) Sokół Orzech
Avoiding direct relegation (10 - 12) Where is My Mind

average value:

11 652 667 zł FC Chełmek
11 517 353 zł KS Polonia Skierni...
11 122 243 zł Patataj
11 059 236 zł KS Wisła Puławy
10 416 364 zł Domik Tuchola
10 228 182 zł Sokół Orzech
10 177 000 zł WHITE TIGERS CF
9 329 758 zł Japanese Futbol
9 212 970 zł ASRIM Iskrzynia
8 285 546 zł Where is My Mind
6 911 862 zł WislaKrakow

How does that promote good tacticians/good transfers/developing young players?

The algorithm does not work

All it does is punishing people with more money on the account
(edited)

That's true. Maybe this is why PSG will never sell out their players to build team only around 17-19yo youngs.

You're missing the point. Nobody cares about POTENTIAL budget in real life. Nobody expected Newcastle United to win Premier League even though it's the most rich team in the league along with Man City, they were expected to finish around 5-8.

Expectations are based on players, history, not on excel spreadsheet.
(edited)

When you will invest your money in young prospects instead of top players - expectations will follow - your nots will go down, your budget wont be to high compared to others - maybe you will get one season of dissapointment from your fanbase, but that is normal, if team that was playing for results shift it strategy.

Your example is perfect - PSG is strong, in real life it is a combination of squad value, budget, history - supporters will expect always first place - even if thay will shift strategy and copy ajax amsterdam. Fans will be dissapointed, and need few season to adjust.

Alternative scenario is pointless - why stats should be the most important thing for expectations? you can easly decrease it in few days selling all players, you can easly improve it when you have lot of money non-used. Stats are important, but overall prepared algoritm is based on differences and many elements, not only stats.

Raul, the best team in the game, playing for 82-84 rating, with many subs/wide rosters, who won last 2 Cups and 2 last Ekstraklasa championships as well as the international Champions Cup is expected to finish at position number NINE in Poland and this will satisfy the supporters.

How can you say that the algorithm is done properly if it works in such a ridiculous manner?

When you will invest your money in young prospects instead of top players - expectations will follow - your nots will go down, your budget wont be to high compared to others - maybe you will get one season of dissapointment from your fanbase, but that is normal, if team that was playing for results shift it strategy.

They apparently don't since two sold out teams from Ekstraklasa playing for 28-36 are expected to finish in Top 3 of the league.

Alternative scenario is pointless - why stats should be the most important thing for expectations?

that's like asking why having best players in the world should be the most important thing for expectations. Because they are the best, strongest etc.
(edited)

I hope nobody will stay under expectation for a long time if he tries to play in a normal way (ie because he is on the same league level for many seasons) - the same like players form, a constant low level kills the fun of your managers, the average should be above the middle.

we introduced it for tests on real database, it is not final, we will analyze it in next few weeks, that is why it is hidden.

They apparently don't since two sold out teams from Ekstraklasa playing for 28-36 are expected to finish in Top 3 of the league.
Yes, expectations are based on last season data, nut current one - and that is good. It would be pointless to adjust expectations with every sell out - it will follow next season.

Damn, once you take team wish has expectations 2-5 and you talk only about beeing 2nd.
Now you take team which has expectations 5-9 and you talk about beeing 9th (why not 5th?). Choosing that border value which fits to your arguments, once highest one, once lowest...

You took average value of the players which also can lie about some statistics.
If I build a 11-man team only from goalkeepers I could have average value even higher that FC Chelmek, this is why you can't take average value to algorithm. But overall value - tells more about money you spent on your team.

Ekstraklasa is not the only league in sokker. But there you can assume that every single user is very good tactician, so their reasults are not based on that factor.

FC Chełmek has best history in last seasons.
They doesnt have highest overall value in league.
Probably they havent got biggest budget. So they don't have biggest possibilities on market to make team stronger.

yes, we think the same way.

Now you take team which has expectations 5-9 and you talk about beeing 9th (why not 5th?)

Because since the best team in the game is EXPECTED to finish in places 5-9 then finishing at 9th is fulfilling the expectations, yes? So the fans of the best team in the game that won numerous trophies in last seasons will be happy if their team finishes 9th. Everything working great :-)

FC Chełmek has best history in last seasons.
They doesnt have highest overall value in league.
Probably they havent got biggest budget. So they don't have biggest possibilities on market to make team stronger.

other team have no chance in reaching their level so what difference it makes? They are the best team, who won the most, why does it matter that they can't get even better if they already are the best/strongest?

that's what I'm saying since the beginning, for the algorithm 1m on the account is worth much more than 1m in players.

team x plays for 84 and has 0m euro in account
team y play for 30 and has 50m euro in account
team y has higher expectations than x

now, team y spends it's money

team x plays for 84 and has 0m in account
team x plays for 75 and has 0m in account
team x has higher expectations than y

it doesn't work, it's just created to bully people with higher transfer balance than those with no money

Same mistakes as in all previous implementations of new features... You guys post it on official server to test it rather than making snapshot of db and then running analysis....

Yes, expectations are based on last season data, nut current one - and that is good.
That's actually very dumb idea :) New player overtaking relegated bot team ain't gonna fulfill those expectations. That player don't know the game yet and will be (in most cases) in league with 2-3 strong opponents.

Also if I understand correctly teams that heavily invested into their relegation/promotion matches will benefit most, because their account balance will decrease (transfers to team for those matches). In the same time those who didn't played those matches and utilized one of the best slots to sell their best players with huge profit, will be expected to win league (bankroll + stats from last season)

@Raul - is there any plan for changing how player's form works? It seems oftentimes some players are stuck in the 0 form for forever

Not forever. Their forms will improve eventually. I was also against the way how the form works here, but now I think it is good how it works and I like the fact that it is generally unknown how to raise player form because if it would be known then everyone would have their playing 11 in top forms which make no sense.

Why very dumb? Even if the bug you mentioned wouldn't going to be fixed, I wouldn't call it "very dumb". It's very negative term, as if everything was wrong, but You point only one very minor issue which is nearly impossible to see and will probably be fixed/adapted in the future, so what else is wrong with it that made you call it "very dumb"?

it is much better place to improve our algoritm, especially if there is no link to any game element.
We want to test it here than on test server, when database is small and not relevant.

No, algoritm is working on data from whole season, not only from its last day/week. So it includes changes of budget level, stats level and any others.

I think someday yes - but it is rather far on our priorities.