主题: I.1

I was second with lvl 13 and 100% youth training and got the 3rd pick though it doesn't matter because I only found 1 A and it looks like mojo has him. Well that was a lot of wasted time

If the lottery really is a lottery, then the team with the best facilities/staff has a 25% chance for #1 pick... so statistically out of the 4 drafts in leagues I and II one of those top teams had to get the top pick!

No... that's wrong. If it were a true lottery, there's a weighted chance to keep whatever position you "should" normally have.

In a true lottery, where teams can move no more than 3 places, in order to determine who drafts 1st overall, teams are each given a certain number of balls in the pot based on their pre-lottery seed. 1st, 2nd, 3rd, and 4th seeds would get 4, 3, 2, and 1 ball(s) respectively when first pick is being determined.

If you do the math, the 1st seeded team would actually have a 40% chance of getting first pick. The 2nd seeded team would have a 30% chance of getting first pick. The chance that either 3rd or 4th seed gets the pick is 30% total.

Let's say that the 3rd seeded team gets it though. The exercise would repeat for the 2nd pick (1st, 2nd, 4th, and 5th would each have 4, 3, 2, and 1 ball(s) in the pot, respectively). So once again... 1st seed has a 40% chance and 2nd seed has a 30% chance. The chance that either 4th or 5th seed gets the pick is 30% total.

The chance that 1st or 2nd seed wouldn't get EITHER of the first two draft picks (what happened in division II.1) is 9%.

OK I see that our definitions of lottery are different. I just meant a simple, unweighted draw. I am unfamiliar with how a "true lottery" works, so I trust you know what you're talking about here. Hopefully this is actually the system PPM uses, since it would be less "random."

I’m too lazy to look up a formal definition, but I would assume that a true lottery would include probabilities greater than 0 for all participants and would not include the possibility of conditional probabilities equal to 1 for any pick besides the last pick (such a case can exist if the PPM draft order is determined one pick at a time). From a mathematical perspective, those requirements would make the lottery more pure and truly random.

The guide states that “..the team with the best academy and sporting directors with the highest youth training efficiency has the biggest chance to pick first.” This rules out the top pick being drawn from a uniform distribution of the top 4 teams. It does not specify the chance for that team (or any other team) to get the top pick. If you have actual data that illustrates those probabilities, then please share. I’m not a fan of un-supported statistics being stated as fact; especially when developed to support an opinion. This is a personal pet peeve of mine and the addition of the phrase “assuming that…” goes a long way in my book.

Having the top sports academy in your league guarantees you a top 4 pick for the mid-season draft each year. Every other week is unaffected. This affects the value of Sports Academies and Directors (slightly), but that knowledge has been openly available.

Having the top sports academy/directors in your league implies that you cannot win (i.e. move up in draft position), but can only tie (pick 1st) or lose (pick 2nd-4th). I believe this contributes to a lot of the complaints people express on the forums. I find this similar to teams complaining about how they are upset more often than they upset other teams, when they’re favored in most (ex: 95%) of their matches.

A benefit of this system is to prevent teams from purposely staying in lower leagues in order to guarantee themselves the top pick in the mid-season draft.

In case you’re wondering, I have the 4th pick in II.3 this draft with 100% youth training.

yes a lottery would allow any team to have a chance of #1 pick.

while id not want managers staying lower in the league just to grab a bunch of great picks Id not want neither rich teams to become rich.

Lets talk about soccer and hockey. I am still struggling to earn money to catch up with facilities. so while i struggle, rich teams get richer with higher picks. My example is for anyone that didn't start on 1st season.

I think a team with lower strength and OTR that does better than stronger/better team should be rewarded. simple as that. so a good team staying on the bottom would not benefit from that. And "bad" teams fighting to get better would be awarded with much better pick.

so the same would work for all those top teams in D1 that started on 1st season. where the weaker performing well would get priority. that way FC Corona, for example, would never get 1st pick allowing others to catch up.


without a good competition it can easily get bored to everyone.

I find it ironic that FC Corona has never picked in the top 10; obviously that wouldn't be the case if drafts had existed in the early years of soccer.

In my opinion, [hockey_team=25869] is the blueprint to starting late in a sport and catching up.

lol, i just mentioned cause i know Corona has a too strong team and to give an example

have to say .. impressive what suprvilce has done. Part of the reason might be cause he focused on one sport. When others came out and I had the chance to start from season 1, i left my teams in the dust. Now i am back managing them.

so giving him as an example, he should be rewarded with great top picks.

so i guess it should be a combination of several things. team strength, OTR and how many days the manager played to get the team to that OTR/Strength

hey... yeah sorry if I annoyed you by putting up "un-supported statistics." I thought it was understood that it was merely an example of how a lottery would work in this situation.

Not only that... but seriously, all of statistics is built on assumptions. You're assuming that your sample size is representative of the population. Assuming that the regression of best fit is the best predictor of future data. Assuming that the variables you've included are the most important. Assuming that the variables that you didn't include are arbitrary.

There's no such thing as statistics that are fact. All statistics are mere approximations. So to assume that I was posting something like that and claiming it to be fact is very presumptive on your part. =/

That's crappy btw that you got stuck with the 4th pick. There's something wrong about the "lottery..." because the chances of that happening should be pretty slim if there's any reasonable amount of weighting in the system.

I'm not going to get into the discussion again about whether the current mid-season draft is fair, but I do want to address one argument that has come up.

The utility of getting the first overall pick with x% is greater than the utility of getting greater than or equal to the 16th pick (as the 16th place team in the rankings) with y%>x%. I don't think it's right to compare 1st place's 0% of moving up with 16th place's z%>0% of moving up and say that 16th place is getting an unfair advantage. While it is disappointing to move back in the draft, the disappointment of moving back alone doesn't affect the ultimate relative outputs from the draft.

Whether the utility, risk aversion, and the probabilities work out to make the draft sufficiently fair to a particular person and whether the probabilities (whatever they are) need to be adjusted to make it fair, I can't say.

Got first in D2 for mine as well. 2nd time I have had 1st pick due to facilities and staff.

This is all absolutely wrong! In a "true" lottery, I go to the lottery winner's house and hold the family hostage until I get the money. Of course, the fatal assumption in this instance is that the lottery winner has a family with which to start...

Again, it's a personal peeve. You may or may not be the one to need to explain to a new player that those numbers are not the actual values (or necessarily close); especially if other players spread those figures.

The draft order has a finite number of variables and the guide would lead you to believe they are sports academy level and youth training efficiency (most likely (facility level)×(1+(eff/2)) - a common formula in the game). The probabilities are either directly programmed into the engine or could be derived directly from the engine (the programmed nature implies finite variables). Unlike more complex examples in the real world, the programmed nature of the logic should result in a perfect match (aiming the formula doesn't change). If you can't replicate the results exactly, then either the inputs are inaccurate or the modeled solution is inaccurate. If I had sufficent data I would probably create a Markov chain from observed probabilities rather than, say, a binomial logit regression. Of course, I enjoy Bayesian probability a bit more than others, so perhaps I'm overly biased.

Also, an example where the team with the best SA/SD having a better chance at getting the 3rd pick than the 4th pick would be more credible (assuming that your example continues with the top team always having a 40% chance to get the 3rd pick if they haven't gotten the 1st or 2nd pick).

I'm not overly concerned with getting the 4th pick. It's something I knew may happen when I invested heavily in the SA. The mid-season draft doesn't have near the importance of the end of season draft (8 weeks of extra training/experience) and I've had a lot of luck with this year's prospects.

Meant to say that if you can't replicate the results within a reasonable confidence interval than the inputs or model are incorrect.