BuzzerBeater Forums

@ Perpete : true.

@ seelenjaeger :
Do correct me if i get you wrong. I'll try to reword your suggestion

A seller will always sell players to the BB-Bank.
The BB-Bank sets the price.
A buyer will always buy from the BB-Bank.
The price is set in auction/bidding scheme (3 steps in starting price, decreasing from 100%-75%-45% then retiring).

The tricky part here is the price of the sell to the BB-Bank. IMO it will be very difficult to set a right price, taking into account all players criteria, and also market needs.
For instance, the major part of managers (please correct if i am wrong) need more allstar->perennial allstar potential players that are well trained and near cap-values than 18 yo HoF. How do the BB-Bank set its price in this case ?
(i imagine that every manager wants to promote or just to improve his results... tanking subject not mentionned)

You´re completely right. The key and crucial part would be to set up a formula for at least some kind of evaluation how much a player is worth.

Sounds like a Mission Impossible at first sight, but there´s TONS of data of recent moves, and once it´s done, it might not be fair from the very point it happened, but it will be in balance in a very short amount of time.

I don´t say I have a formula about this, and I´m sure we all know the current TPA is not handy for that, but I´m sure it can be done and the benefit will outweigh the momentary trouble situation when introducing the system. Could be a step-by-step introduction by the way. Keep the "sell to the bank or via auction" option for 2 seasons so people get used to it, and once they are, take away the direct sell to another player and make it bank only. Mathematical systems are able to adopt quickly using big data amounts.

I find an 18 year old guy with MVP potential on the list for 300k ... perfect size ... train him for 4-5 weeks so that people start realizing where he might end up at. I resell him for 1 - 1.2 Mios, as he might easily be worth it. Where is the part when I did something wrong that justifies me ending up with 400k only?

The only thing i'd like to point out here, is that the data might be "corrupted" by many current and past "unfair" trades (i don't say that it is the case because i'm not sure about the rates).
As i write those lines, i must admit that it seems quite a subjective issue, what is a "fair" price and what is not.

As a complement to your suggestion, i will gladly add some kind of adjustment linked to the demand/offer. For example, if there are 1000 sells (from managers to BB-Bank) at the same period for similar kind of players, the BB-Bank adjusts prices decreasing. And vice-versa if there are very few sells, the price set by the BB-Bank is at the max value.
Still, how to set up this decreasing formula ... ? Exponential ? Linear ? With a minimal cap ?

I must confess that I didnt read each and any point in the current discussion, but I come to the conclusion that you try to build up an immense rulework to cure a limited problem.

Hummm ... how can you come to a conclusion about someone's suggestion without hearing all about it ? :°)

My thoughts exactly.

Well, first of all - I think there´s two different discussions.

(1) IF there is a working formular, would the proposed system be desireable?
and
(2) Is it somehow possible to find such a formula?

About 1: If there would be any way to implement a seperation between buyer and seller, the benefit would easily outweigh the downside from my point of view.

About 2: I do think that based on the collected data over the seasons, the existing formulae which create salary, rating, efficiency and the information about how the skills and age and size affect each other when performing training (including the hidden skills) should provide such a heavy data base that it should be quite easy to find a correlation between those factors and the paid prices. The mathematical genious needed to program a working game engine (or overworking it) is probably alot bigger than whatever is needed to describe an existing process giving all the influencal insight the BBs have.

Just take a look at JosefKas work in deciphering many aspects of the GE - he had just a small amount of the overall data and still managed to set up models with pretty distinct and precise information about the outcome. Given, he was giving projections about mechanisms, not user behaviour, but the user behaviour is the very thing everybody wants to elminate when talking about fair prices.

By monitoring the "buying market", which would still be independet from the selling market, you also get a good feeling for changes in preferences, style, importance, so that any "market value" formular can hold a dynamic factor taking into account how "sexy" players are, for example - just as Wolph mentioned in another context - when beeing injured 3, 4, 5 weeks, rooks are highly interesting at the beginning of a season, lose value by the end, veterans gain value come playoff time, and so on.

Setting up such a system might be quite a workload, but with mathematical skill and a solid model it would probably be a huge step towards Buzzerbeater 2.0

(1) IF there is a working formular, would the proposed system be desireable?
and
(2) Is it somehow possible to find such a formula?

I guess you already pretty much answered this way in your post, but I think it is worth repeating.

(1) Perhaps so... From the "reality" point of view (and most of us probably consider that to some degree about most everything in this game), it would seem weird and wrong. From the game-design point of view, it might work all right. The buying decision is about improving your team. The selling price is currently somewhat up to luck. Both are affected by the market, but not always to same extent (we all know the complaints about time of auction deadlines etc.)

(2) Considering that implementing the formula would remove much of incentive for overpaying (current TPE considers also overpriced transfers), once the change is in place the market data would certainly be better than what we have today for TPE.

Your suggestion seems quite efficient to me as it will make huge changes. Still, it requires a solid analysis of the data to make coherence with other existing criteria of the players.

I'll be short. The following figures about any player are already available and bring no interpretation :
- last trading date
- last trading price.

With an easy calculation, you can come out with how long the player was on a roster. You link that to a profit tax.
1 - 2 weeks : 100% profit tax (no profit made on the trade)
3 - 5 weeks : 75% profit tax (25% maximum profit made by seller)
6 - 9 weeks : 50% profit tax (50% max profit )
10 - 14 weeks : 25% profit tax (75% max profit made)
> 1 season : No profit tax.

Of course, current trade fees disappear.
Nothing else changes from current Trade scheme (bidding, setting sales price and how long the player is on TL).
You can imagine the player going to FA if no one bids. Could be complementary.

Seems quite simplier... And guess what, that's what Wolph was all about.

PS.: i divided 1 season time into 4 growing parts (2, 3, 4 and 5 weeks long) and accord them +25% of profit tax. Too simple to be fair enough, consider it as a sketch that can/must be sharpenned, but you got the main idea of it.