BuzzerBeater Forums

of the top 33 most populated countries in buzzerbeater (country 33 has 414 users), meaning 33*4=132 division 1 promotions, there were 60 subsequent demotions, making it a 55% survival rate, which i agree is lower... but also is just about where you thought it should be, and where we think it should be too.

fwiw.. in the top 10 countries, it was 16 demotions.. or 60% survival rate.

Ok then I missed that as we were posting simultanouesly. I still think there is an unreasonable effort to make gaps in the future.)

Another thing to analyse could how much those Divs actaully changes and how much teams were "always there" like in last 5 season in a row. But I'm not asking you to do that.

I could look at that.. but actually, if your model of what is correct is 4 random teams are demoted every season that means either parity if you are an optimist, or means there is no effect of manager skill on the game.

meaning if there is a lot of turnover, how could it be possible that someone is actually "good" at this game in any meaningful way?

I meant it more liek probability to actually catch up with top team of the country.

We are looking at numbers in our country and have a bit different resultzs I guess it is a local thing then.

It seems that my point about the big8 vs great8 was missing during this big "conversation".

I still would like to see some reaction on that: (136516.170)

but your country only has 4 teams promote to d1 every season...

so if you work out the statistics, if you thought each team had a 60% chance of demoting

2.56 % of the time you would see all 4 demote
15.36 % of the time you would see 3 of 4 demote
34.56 % of the time you would see 2 of 4 demote
34.56 % of the time you would see 1 of 4 demote
12.96 % of the time you would see 0 of 4 demote

So really... none of those possibilities are really that crazy to imagine... except maybe all 4 going down has a p value of less than .05, the weakest test science would accept as significant.

Maybe it's really local think for Czech republic, but it really looks hard for promoted teams to establish in Div I. We have in our Div I - 9 team playing Div I for 5 of more seasons in a row and 3 other teams which missed only one season for last 5 seasons, so it's 12 very well established team and 1 more which demoted last season after 9 seasons in Div I.

Season 11 - 3 of 4 demoted
Season 10 - 2 of 4 demoted
Season 9 - 3 of 4 demoted
Season 8 - 3 of 4 demoted
Season 7 - 2 of 5 demoted

It looks our country find some equlibrum since season 8 and it's really hard to fight against well-settled teams.

Last edited by rwystyrk at 3/29/2010 12:03:57 PM

I propose that every season the teams of the same division are shuffled and you get new conferences every season.

I think this will not happen, because during the seasons are teams making rivals between themselves and other tiny things which is "you were above me last time" are creating something like conference character. The idea you propose would ruin everything.

So running the numbers on that, it looks to be about a 1.5 sigma effect, or you'd expect fewer teams to promote only 3.25% of the time... so that's approaching significant.

Of course you can argue that given that there are ~30-100 countries in the game, you should expect to see an event like this somewhere in buzzerbeater.. 1/100 being 1%, 1/30 being 3.3%. So in that context its not that unexpected.

Its also the case that such an analysis doesn't really take into account local variations.. like the possibility that this group of 9 is actually much better at the game and therefore deserves to be in perennially.

All in all the difference between the most likely result 12-13 teams survive, versus 8, is only 4 teams over those 5 seasons. So its not really a big difference.

Last edited by BB-Forrest at 3/29/2010 1:14:47 PM

So you prefer to have competition between your conference than competition between conferences!

I understand what you mean by character but the team will not change division, only conference!

Plus, I think that conferences where you have the champion will have more probability of receiving someone from the upper division which means they will still be stronger (if we assume that there is a high chance of the relegated team to promote again). But because I can't tell this without numbers, I had a look into the Portuguese division III over the last 6 seasons and check whether we have continuous champions from the same conference in all the divisions.

This are the results (R for red conference and B for blue conference)

League 1 - R-R-B-R-R-R     (83% Red)
League 2 - R-B-R-R-R-R     (83% Red)
League 3 - B-R-R-B-R-R     (67% Red)
League 4 - R-B-R-R-R-R     (83% Red)
League 5 - R-R-R-B-B-R     (67% Red)
League 6 - B-R-B-R-R-B     (50% Red)
League 7 - B-B-B-R-B-B     (83% Blue)
League 8 - R-R-R-R-R-B     (83% Red)
League 9 - R-R-R-B-B-R     (67% Red)
League 10 - R-R-R-R-R-R     (100% Red)
League 11 - R-B-R-R-R-R     (83% Red)
League 12 - B-R-B-R-B-B     (50% Red)
League 13 - R-B-B-R-R-R     (67% Red)
League 14 - R-B-B-B-R-R     (50% Red)
League 15 - B-R-R-R-R-R     (83% Red)
League 16 - R-B-B-R-R-B     (50% Red)

A quick look will tell you that only 4 leagues out of 16 had half of his champions from both conferences. 7 out of 16 (almost 50%) leagues had 5 times a champion from one of the conferences in 6 seasons. And... just because it is funny to mention, 64 out of the 96 (67%) champions were from the Red conference!

Another interesting fact is that most of the times, the champion team is from the same conference of the last champion of that same division (aproximatly 70% of the times)!

Of course this is really a small sample but I do believe it could be the same in most of the countries. Or maybe I am wrong and Portugal just likes Red to much (Benfica xD)

Last edited by Killer Penguin at 3/29/2010 1:38:25 PM

did you know the expected number is 3.5 of those leagues should have a champion from one of the conferences 5 or more times? So really if just 3 championship games turned out a different result it would be the expected number of events?

I know it sounds unlikely when you say one side won 5 out of 6 times, but you are counting it also unlikely when the other side wins 5 out of 6... and really when you are flipping coins... 5 out of 6 heads or tails really isn't that unlikely. 11% of the time you get at least 5 out of 6 heads, and since you are counting both 5/6 heads and 5/6 tails, its really 22%.

Look,if you want to shuffle them every season, then would be better to not have conferences at all and just have 1 table like in football league.

When I saw this conference system at first, I was happy to have it here, because it even create an opportunity to focus more on just few key teams to beat down and the playoff is yours.

Your red paranoea is based just on that there has to be probability that it should be like 50/50 in overall but it is not. Well I can tell you that teams are not just a numbers, I would be very suprised if there would be the equilibrium. I can understand that it sounds scarry if you look at it by this way, but these things happens, it is just a collor, thats all.